WEBVTT Kind: captions Language: en 00:00:01.160 --> 00:00:06.690 align:start position:0% [Music] 00:00:06.690 --> 00:00:06.700 align:start position:0% 00:00:06.700 --> 00:00:10.430 align:start position:0% [Music] 00:00:10.430 --> 00:00:10.440 align:start position:0% [Music] 00:00:10.440 --> 00:00:12.990 align:start position:0% [Music] uh<00:00:11.440> hello<00:00:11.799> guys<00:00:12.280> uh<00:00:12.440> thank<00:00:12.559> you<00:00:12.639> for<00:00:12.799> coming 00:00:12.990 --> 00:00:13.000 align:start position:0% uh hello guys uh thank you for coming 00:00:13.000 --> 00:00:13.829 align:start position:0% uh hello guys uh thank you for coming for<00:00:13.120> the 00:00:13.829 --> 00:00:13.839 align:start position:0% for the 00:00:13.839 --> 00:00:16.830 align:start position:0% for the presentation<00:00:14.839> um<00:00:15.519> okay<00:00:15.799> I<00:00:15.920> think<00:00:16.080> I<00:00:16.199> Buist 00:00:16.830 --> 00:00:16.840 align:start position:0% presentation um okay I think I Buist 00:00:16.840 --> 00:00:19.070 align:start position:0% presentation um okay I think I Buist click<00:00:17.480> so<00:00:17.840> uh<00:00:17.960> we<00:00:18.039> are<00:00:18.240> presenting<00:00:18.600> the<00:00:18.760> rated 00:00:19.070 --> 00:00:19.080 align:start position:0% click so uh we are presenting the rated 00:00:19.080 --> 00:00:22.470 align:start position:0% click so uh we are presenting the rated list<00:00:19.840> uh<00:00:20.000> we<00:00:20.119> were<00:00:20.240> mentored<00:00:20.680> by<00:00:20.960> danrad<00:00:21.960> uh<00:00:22.320> on 00:00:22.470 --> 00:00:22.480 align:start position:0% list uh we were mentored by danrad uh on 00:00:22.480 --> 00:00:25.070 align:start position:0% list uh we were mentored by danrad uh on this<00:00:22.680> project<00:00:23.000> it<00:00:23.119> was<00:00:23.320> his<00:00:23.480> idea<00:00:24.080> originally 00:00:25.070 --> 00:00:25.080 align:start position:0% this project it was his idea originally 00:00:25.080 --> 00:00:27.790 align:start position:0% this project it was his idea originally uh<00:00:25.320> what<00:00:25.519> we<00:00:25.720> did<00:00:26.080> was<00:00:26.800> uh<00:00:26.960> to<00:00:27.279> formally 00:00:27.790 --> 00:00:27.800 align:start position:0% uh what we did was uh to formally 00:00:27.800 --> 00:00:30.429 align:start position:0% uh what we did was uh to formally specify<00:00:28.359> the<00:00:28.480> rated<00:00:28.840> list<00:00:29.720> uh<00:00:30.000> take<00:00:30.160> it<00:00:30.279> out 00:00:30.429 --> 00:00:30.439 align:start position:0% specify the rated list uh take it out 00:00:30.439 --> 00:00:34.430 align:start position:0% specify the rated list uh take it out from<00:00:30.560> a<00:00:30.720> Blog<00:00:31.279> and<00:00:31.560> have<00:00:31.720> a<00:00:31.920> proper<00:00:32.680> uh<00:00:33.480> set<00:00:33.760> of 00:00:34.430 --> 00:00:34.440 align:start position:0% from a Blog and have a proper uh set of 00:00:34.440 --> 00:00:37.990 align:start position:0% from a Blog and have a proper uh set of uh<00:00:34.960> pythonic<00:00:35.640> specs<00:00:36.600> uh<00:00:36.680> for<00:00:36.879> the<00:00:37.000> rated<00:00:37.360> list 00:00:37.990 --> 00:00:38.000 align:start position:0% uh pythonic specs uh for the rated list 00:00:38.000 --> 00:00:42.029 align:start position:0% uh pythonic specs uh for the rated list build<00:00:38.200> a<00:00:38.520> simulator<00:00:39.520> uh<00:00:40.320> which<00:00:41.320> against<00:00:41.800> which 00:00:42.029 --> 00:00:42.039 align:start position:0% build a simulator uh which against which 00:00:42.039 --> 00:00:44.549 align:start position:0% build a simulator uh which against which we<00:00:42.160> can<00:00:42.600> uh<00:00:42.719> test<00:00:43.000> the<00:00:43.079> rated<00:00:43.480> list<00:00:44.239> uh<00:00:44.360> write 00:00:44.549 --> 00:00:44.559 align:start position:0% we can uh test the rated list uh write 00:00:44.559 --> 00:00:46.709 align:start position:0% we can uh test the rated list uh write some<00:00:44.719> unit<00:00:45.039> tests<00:00:45.640> so<00:00:45.840> that<00:00:46.079> we<00:00:46.239> have<00:00:46.480> some 00:00:46.709 --> 00:00:46.719 align:start position:0% some unit tests so that we have some 00:00:46.719 --> 00:00:49.110 align:start position:0% some unit tests so that we have some basic<00:00:47.079> sanity<00:00:47.520> checks<00:00:48.000> of<00:00:48.199> the<00:00:48.320> rated<00:00:48.640> list<00:00:48.960> so 00:00:49.110 --> 00:00:49.120 align:start position:0% basic sanity checks of the rated list so 00:00:49.120 --> 00:00:51.350 align:start position:0% basic sanity checks of the rated list so that<00:00:49.280> it<00:00:49.399> does<00:00:49.559> not<00:00:49.800> perform<00:00:50.680> worser<00:00:51.160> than 00:00:51.350 --> 00:00:51.360 align:start position:0% that it does not perform worser than 00:00:51.360 --> 00:00:53.310 align:start position:0% that it does not perform worser than just<00:00:51.600> randomly<00:00:52.039> pure 00:00:53.310 --> 00:00:53.320 align:start position:0% just randomly pure 00:00:53.320 --> 00:00:55.830 align:start position:0% just randomly pure sampling<00:00:54.320> uh<00:00:54.600> and<00:00:54.879> collect<00:00:55.199> some<00:00:55.359> metrics 00:00:55.830 --> 00:00:55.840 align:start position:0% sampling uh and collect some metrics 00:00:55.840 --> 00:00:59.069 align:start position:0% sampling uh and collect some metrics against<00:00:56.640> uh<00:00:57.000> known<00:00:57.359> attacks<00:00:58.359> uh<00:00:58.480> on<00:00:58.719> existing 00:00:59.069 --> 00:00:59.079 align:start position:0% against uh known attacks uh on existing 00:00:59.079 --> 00:01:03.430 align:start position:0% against uh known attacks uh on existing dhds<00:01:00.199> so<00:01:00.800> uh<00:01:01.000> just<00:01:01.120> to<00:01:01.359> give<00:01:01.519> you<00:01:02.199> a<00:01:03.199> quick 00:01:03.430 --> 00:01:03.440 align:start position:0% dhds so uh just to give you a quick 00:01:03.440 --> 00:01:05.990 align:start position:0% dhds so uh just to give you a quick intro<00:01:03.800> to<00:01:03.960> rated<00:01:04.479> list 00:01:05.990 --> 00:01:06.000 align:start position:0% intro to rated list 00:01:06.000 --> 00:01:08.870 align:start position:0% intro to rated list sorry<00:01:07.000> um<00:01:07.280> so<00:01:07.520> this<00:01:07.640> was<00:01:07.960> the<00:01:08.159> idea<00:01:08.479> proposed 00:01:08.870 --> 00:01:08.880 align:start position:0% sorry um so this was the idea proposed 00:01:08.880 --> 00:01:13.109 align:start position:0% sorry um so this was the idea proposed by<00:01:09.479> danrad<00:01:10.479> um<00:01:11.400> which<00:01:11.560> is<00:01:12.159> uh<00:01:12.360> instead<00:01:12.720> of 00:01:13.109 --> 00:01:13.119 align:start position:0% by danrad um which is uh instead of 00:01:13.119 --> 00:01:16.390 align:start position:0% by danrad um which is uh instead of having<00:01:13.320> a<00:01:13.479> DST<00:01:14.400> uh<00:01:14.560> you<00:01:14.759> form<00:01:15.200> a<00:01:15.560> tree<00:01:16.040> of<00:01:16.240> all 00:01:16.390 --> 00:01:16.400 align:start position:0% having a DST uh you form a tree of all 00:01:16.400 --> 00:01:19.230 align:start position:0% having a DST uh you form a tree of all the<00:01:16.520> nodes<00:01:16.960> that<00:01:17.080> you<00:01:17.240> know<00:01:17.880> you<00:01:18.159> ask<00:01:18.439> for<00:01:19.040> uh 00:01:19.230 --> 00:01:19.240 align:start position:0% the nodes that you know you ask for uh 00:01:19.240 --> 00:01:20.990 align:start position:0% the nodes that you know you ask for uh in<00:01:19.439> kind<00:01:19.560> of<00:01:19.680> a<00:01:19.799> pure<00:01:20.079> exchange<00:01:20.479> manner<00:01:20.840> you 00:01:20.990 --> 00:01:21.000 align:start position:0% in kind of a pure exchange manner you 00:01:21.000 --> 00:01:23.670 align:start position:0% in kind of a pure exchange manner you ask<00:01:21.200> for<00:01:21.400> nodes<00:01:22.320> of<00:01:22.520> the<00:01:22.640> nodes<00:01:23.240> and<00:01:23.320> you<00:01:23.439> build 00:01:23.670 --> 00:01:23.680 align:start position:0% ask for nodes of the nodes and you build 00:01:23.680 --> 00:01:26.670 align:start position:0% ask for nodes of the nodes and you build out<00:01:23.799> a<00:01:24.000> tree<00:01:24.720> and<00:01:25.439> uh<00:01:25.880> you<00:01:26.040> start<00:01:26.280> your<00:01:26.400> peer 00:01:26.670 --> 00:01:26.680 align:start position:0% out a tree and uh you start your peer 00:01:26.680 --> 00:01:29.390 align:start position:0% out a tree and uh you start your peer sampling<00:01:27.400> and<00:01:27.640> based<00:01:27.880> on<00:01:28.000> the<00:01:28.119> peer<00:01:28.400> sampling 00:01:29.390 --> 00:01:29.400 align:start position:0% sampling and based on the peer sampling 00:01:29.400 --> 00:01:32.710 align:start position:0% sampling and based on the peer sampling uh<00:01:29.520> if<00:01:29.720> a<00:01:29.960> node<00:01:30.240> replies<00:01:30.840> or<00:01:31.000> does<00:01:31.200> not<00:01:31.720> reply 00:01:32.710 --> 00:01:32.720 align:start position:0% uh if a node replies or does not reply 00:01:32.720 --> 00:01:35.389 align:start position:0% uh if a node replies or does not reply uh<00:01:33.159> you<00:01:33.479> propagate<00:01:34.040> those<00:01:34.240> scores<00:01:35.000> back<00:01:35.200> to 00:01:35.389 --> 00:01:35.399 align:start position:0% uh you propagate those scores back to 00:01:35.399 --> 00:01:38.310 align:start position:0% uh you propagate those scores back to all<00:01:35.600> the<00:01:36.040> ancestors<00:01:37.040> right<00:01:37.479> so<00:01:38.000> as<00:01:38.119> you<00:01:38.200> can 00:01:38.310 --> 00:01:38.320 align:start position:0% all the ancestors right so as you can 00:01:38.320 --> 00:01:41.030 align:start position:0% all the ancestors right so as you can see<00:01:38.560> the<00:01:38.680> red<00:01:38.920> node<00:01:39.399> is<00:01:39.759> a<00:01:40.119> malicious<00:01:40.640> node<00:01:40.880> or 00:01:41.030 --> 00:01:41.040 align:start position:0% see the red node is a malicious node or 00:01:41.040 --> 00:01:43.030 align:start position:0% see the red node is a malicious node or a<00:01:41.159> node<00:01:41.439> that<00:01:41.560> does<00:01:41.720> not<00:01:41.920> respond<00:01:42.640> and<00:01:42.799> because 00:01:43.030 --> 00:01:43.040 align:start position:0% a node that does not respond and because 00:01:43.040 --> 00:01:46.069 align:start position:0% a node that does not respond and because of<00:01:43.320> that<00:01:43.799> the<00:01:43.960> scores<00:01:44.360> of<00:01:44.719> its<00:01:44.960> direct<00:01:45.320> parent 00:01:46.069 --> 00:01:46.079 align:start position:0% of that the scores of its direct parent 00:01:46.079 --> 00:01:48.590 align:start position:0% of that the scores of its direct parent and<00:01:46.320> its<00:01:46.600> Grand<00:01:46.960> great<00:01:47.240> grandparents<00:01:48.079> are 00:01:48.590 --> 00:01:48.600 align:start position:0% and its Grand great grandparents are 00:01:48.600 --> 00:01:50.910 align:start position:0% and its Grand great grandparents are affected<00:01:49.600> right<00:01:49.840> so<00:01:50.079> this<00:01:50.200> is<00:01:50.320> the<00:01:50.439> basic<00:01:50.719> idea 00:01:50.910 --> 00:01:50.920 align:start position:0% affected right so this is the basic idea 00:01:50.920 --> 00:01:53.910 align:start position:0% affected right so this is the basic idea of<00:01:51.040> the<00:01:51.119> rated<00:01:51.560> list<00:01:52.560> uh<00:01:53.200> this<00:01:53.360> was<00:01:53.560> what<00:01:53.680> was 00:01:53.910 --> 00:01:53.920 align:start position:0% of the rated list uh this was what was 00:01:53.920 --> 00:01:56.190 align:start position:0% of the rated list uh this was what was defined<00:01:54.360> in<00:01:54.479> the<00:01:54.600> blog<00:01:55.360> uh<00:01:55.479> we<00:01:55.640> went<00:01:55.840> on<00:01:56.039> to 00:01:56.190 --> 00:01:56.200 align:start position:0% defined in the blog uh we went on to 00:01:56.200 --> 00:01:57.590 align:start position:0% defined in the blog uh we went on to further<00:01:56.560> Define<00:01:56.960> how<00:01:57.119> you<00:01:57.240> would<00:01:57.399> use 00:01:57.590 --> 00:01:57.600 align:start position:0% further Define how you would use 00:01:57.600 --> 00:02:02.469 align:start position:0% further Define how you would use filtering<00:01:58.039> over<00:01:58.280> it<00:01:59.159> uh<00:01:59.600> which<00:02:00.159> is<00:02:01.200> uh<00:02:02.200> sorry 00:02:02.469 --> 00:02:02.479 align:start position:0% filtering over it uh which is uh sorry 00:02:02.479 --> 00:02:05.109 align:start position:0% filtering over it uh which is uh sorry scoring<00:02:03.119> yeah<00:02:03.520> so<00:02:03.759> once<00:02:03.960> you<00:02:04.159> have<00:02:04.759> uh<00:02:04.920> every 00:02:05.109 --> 00:02:05.119 align:start position:0% scoring yeah so once you have uh every 00:02:05.119 --> 00:02:08.669 align:start position:0% scoring yeah so once you have uh every node<00:02:05.560> descendant<00:02:06.039> score<00:02:06.880> uh<00:02:07.039> so<00:02:07.320> if<00:02:07.920> a<00:02:08.080> set<00:02:08.239> of 00:02:08.669 --> 00:02:08.679 align:start position:0% node descendant score uh so if a set of 00:02:08.679 --> 00:02:12.070 align:start position:0% node descendant score uh so if a set of uh<00:02:08.800> nodes<00:02:09.160> are<00:02:09.319> serving<00:02:09.679> a<00:02:10.160> sample<00:02:11.160> uh<00:02:11.840> their 00:02:12.070 --> 00:02:12.080 align:start position:0% uh nodes are serving a sample uh their 00:02:12.080 --> 00:02:14.830 align:start position:0% uh nodes are serving a sample uh their descendant<00:02:12.720> scores<00:02:13.720> are<00:02:14.080> not<00:02:14.319> the<00:02:14.560> actual 00:02:14.830 --> 00:02:14.840 align:start position:0% descendant scores are not the actual 00:02:14.840 --> 00:02:16.670 align:start position:0% descendant scores are not the actual scores<00:02:15.160> that<00:02:15.280> we<00:02:15.360> use<00:02:15.560> for<00:02:15.720> filtering<00:02:16.319> what<00:02:16.480> we 00:02:16.670 --> 00:02:16.680 align:start position:0% scores that we use for filtering what we 00:02:16.680 --> 00:02:19.110 align:start position:0% scores that we use for filtering what we do<00:02:17.000> is<00:02:17.760> uh<00:02:18.040> we 00:02:19.110 --> 00:02:19.120 align:start position:0% do is uh we 00:02:19.120 --> 00:02:22.990 align:start position:0% do is uh we trickle<00:02:20.120> uh<00:02:20.440> not<00:02:20.640> trickle<00:02:21.160> basically<00:02:21.599> but<00:02:22.160> um 00:02:22.990 --> 00:02:23.000 align:start position:0% trickle uh not trickle basically but um 00:02:23.000 --> 00:02:26.070 align:start position:0% trickle uh not trickle basically but um we<00:02:23.200> kind<00:02:23.400> of<00:02:23.519> follow<00:02:23.959> the<00:02:24.200> paths<00:02:25.120> of<00:02:25.680> uh<00:02:25.800> that 00:02:26.070 --> 00:02:26.080 align:start position:0% we kind of follow the paths of uh that 00:02:26.080 --> 00:02:28.589 align:start position:0% we kind of follow the paths of uh that node<00:02:26.400> existence<00:02:27.239> in<00:02:27.560> all<00:02:27.760> the<00:02:27.879> sub<00:02:28.160> trees 00:02:28.589 --> 00:02:28.599 align:start position:0% node existence in all the sub trees 00:02:28.599 --> 00:02:30.869 align:start position:0% node existence in all the sub trees because<00:02:29.239> a<00:02:29.400> node<00:02:29.640> has<00:02:30.040> many<00:02:30.239> peers<00:02:30.640> and<00:02:30.760> it 00:02:30.869 --> 00:02:30.879 align:start position:0% because a node has many peers and it 00:02:30.879 --> 00:02:33.270 align:start position:0% because a node has many peers and it would<00:02:31.080> exist<00:02:31.319> in<00:02:31.480> different<00:02:31.720> sub<00:02:31.959> trees<00:02:32.840> uh<00:02:33.000> we 00:02:33.270 --> 00:02:33.280 align:start position:0% would exist in different sub trees uh we 00:02:33.280 --> 00:02:36.990 align:start position:0% would exist in different sub trees uh we take<00:02:34.120> the<00:02:34.440> level<00:02:34.720> one<00:02:35.000> peers<00:02:35.879> scores<00:02:36.879> which 00:02:36.990 --> 00:02:37.000 align:start position:0% take the level one peers scores which 00:02:37.000 --> 00:02:39.270 align:start position:0% take the level one peers scores which are<00:02:37.200> the<00:02:37.400> best<00:02:37.680> for<00:02:38.080> that<00:02:38.200> node's<00:02:38.640> path<00:02:39.120> right 00:02:39.270 --> 00:02:39.280 align:start position:0% are the best for that node's path right 00:02:39.280 --> 00:02:41.550 align:start position:0% are the best for that node's path right so<00:02:39.400> it's<00:02:39.519> the<00:02:39.640> best<00:02:39.840> path<00:02:40.080> score<00:02:40.319> of<00:02:40.480> that<00:02:40.599> node 00:02:41.550 --> 00:02:41.560 align:start position:0% so it's the best path score of that node 00:02:41.560 --> 00:02:43.790 align:start position:0% so it's the best path score of that node uh<00:02:41.720> we<00:02:41.879> use<00:02:42.200> that<00:02:42.360> score<00:02:42.920> we<00:02:43.040> filter<00:02:43.400> out<00:02:43.560> using 00:02:43.790 --> 00:02:43.800 align:start position:0% uh we use that score we filter out using 00:02:43.800 --> 00:02:45.990 align:start position:0% uh we use that score we filter out using a<00:02:43.920> threshold<00:02:44.800> if<00:02:44.920> the<00:02:45.080> threshold<00:02:45.519> does<00:02:45.680> not 00:02:45.990 --> 00:02:46.000 align:start position:0% a threshold if the threshold does not 00:02:46.000 --> 00:02:47.589 align:start position:0% a threshold if the threshold does not work<00:02:46.280> like<00:02:46.480> if<00:02:46.560> it<00:02:46.680> does<00:02:46.840> not<00:02:47.000> filter<00:02:47.280> out<00:02:47.440> any 00:02:47.589 --> 00:02:47.599 align:start position:0% work like if it does not filter out any 00:02:47.599 --> 00:02:50.750 align:start position:0% work like if it does not filter out any nodes<00:02:48.360> we<00:02:48.599> switch<00:02:48.920> to<00:02:49.120> an<00:02:49.319> average<00:02:49.720> score<00:02:50.560> uh 00:02:50.750 --> 00:02:50.760 align:start position:0% nodes we switch to an average score uh 00:02:50.760 --> 00:02:54.670 align:start position:0% nodes we switch to an average score uh for<00:02:51.000> the<00:02:51.120> same<00:02:52.080> and<00:02:52.920> um<00:02:53.319> basically<00:02:53.720> because<00:02:54.560> we 00:02:54.670 --> 00:02:54.680 align:start position:0% for the same and um basically because we 00:02:54.680 --> 00:02:57.589 align:start position:0% for the same and um basically because we want<00:02:54.879> to<00:02:55.080> complete<00:02:55.480> sampling<00:02:56.239> even<00:02:56.560> though<00:02:57.239> uh 00:02:57.589 --> 00:02:57.599 align:start position:0% want to complete sampling even though uh 00:02:57.599 --> 00:02:59.229 align:start position:0% want to complete sampling even though uh you<00:02:57.720> know<00:02:58.000> we<00:02:58.120> cannot<00:02:58.400> filter<00:02:58.720> out<00:02:58.879> using<00:02:59.120> the 00:02:59.229 --> 00:02:59.239 align:start position:0% you know we cannot filter out using the 00:02:59.239 --> 00:03:02.869 align:start position:0% you know we cannot filter out using the rated<00:02:59.560> list<00:03:00.319> so<00:03:00.920> that's<00:03:01.720> um<00:03:02.360> that's<00:03:02.560> average 00:03:02.869 --> 00:03:02.879 align:start position:0% rated list so that's um that's average 00:03:02.879 --> 00:03:05.309 align:start position:0% rated list so that's um that's average filtering<00:03:03.799> and<00:03:04.319> after<00:03:04.560> filtering<00:03:05.040> we<00:03:05.159> can 00:03:05.309 --> 00:03:05.319 align:start position:0% filtering and after filtering we can 00:03:05.319 --> 00:03:07.030 align:start position:0% filtering and after filtering we can apply<00:03:05.640> different<00:03:05.920> quering<00:03:06.280> strategies<00:03:06.879> which 00:03:07.030 --> 00:03:07.040 align:start position:0% apply different quering strategies which 00:03:07.040 --> 00:03:09.309 align:start position:0% apply different quering strategies which is<00:03:07.200> we<00:03:07.319> can<00:03:07.480> use<00:03:07.799> a<00:03:07.920> random<00:03:08.200> quering<00:03:08.560> strategy 00:03:09.309 --> 00:03:09.319 align:start position:0% is we can use a random quering strategy 00:03:09.319 --> 00:03:10.949 align:start position:0% is we can use a random quering strategy we<00:03:09.440> can<00:03:09.560> use<00:03:09.760> the<00:03:09.879> highest<00:03:10.159> score<00:03:10.480> first<00:03:10.840> we 00:03:10.949 --> 00:03:10.959 align:start position:0% we can use the highest score first we 00:03:10.959 --> 00:03:12.789 align:start position:0% we can use the highest score first we can<00:03:11.120> use<00:03:11.319> the<00:03:11.400> lowest<00:03:11.680> score<00:03:11.959> first<00:03:12.560> and<00:03:12.680> we 00:03:12.789 --> 00:03:12.799 align:start position:0% can use the lowest score first and we 00:03:12.799 --> 00:03:14.710 align:start position:0% can use the lowest score first and we can<00:03:12.920> use<00:03:13.120> the<00:03:13.239> average<00:03:13.519> score<00:03:13.799> first<00:03:14.400> average 00:03:14.710 --> 00:03:14.720 align:start position:0% can use the average score first average 00:03:14.720 --> 00:03:16.390 align:start position:0% can use the average score first average score<00:03:15.000> first<00:03:15.239> and<00:03:15.360> lowest<00:03:15.640> score<00:03:15.920> first<00:03:16.200> does 00:03:16.390 --> 00:03:16.400 align:start position:0% score first and lowest score first does 00:03:16.400 --> 00:03:18.630 align:start position:0% score first and lowest score first does not<00:03:16.560> make<00:03:16.720> a<00:03:16.840> lot<00:03:16.959> of<00:03:17.120> sense<00:03:17.440> in<00:03:17.560> the<00:03:17.720> start<00:03:18.440> but 00:03:18.630 --> 00:03:18.640 align:start position:0% not make a lot of sense in the start but 00:03:18.640 --> 00:03:21.990 align:start position:0% not make a lot of sense in the start but if<00:03:18.760> you<00:03:18.920> think<00:03:19.159> about<00:03:19.400> it<00:03:20.280> uh<00:03:21.000> parent<00:03:21.400> note<00:03:21.720> can 00:03:21.990 --> 00:03:22.000 align:start position:0% if you think about it uh parent note can 00:03:22.000 --> 00:03:25.869 align:start position:0% if you think about it uh parent note can have<00:03:22.720> an<00:03:23.480> a<00:03:23.640> perfect<00:03:24.360> uh<00:03:24.480> score<00:03:25.360> because<00:03:25.680> none 00:03:25.869 --> 00:03:25.879 align:start position:0% have an a perfect uh score because none 00:03:25.879 --> 00:03:28.589 align:start position:0% have an a perfect uh score because none of<00:03:26.040> its<00:03:26.319> children<00:03:26.640> are<00:03:26.879> contacted<00:03:27.480> yet<00:03:28.040> right 00:03:28.589 --> 00:03:28.599 align:start position:0% of its children are contacted yet right 00:03:28.599 --> 00:03:31.630 align:start position:0% of its children are contacted yet right so<00:03:29.159> uh<00:03:29.280> highest<00:03:29.560> score<00:03:29.920> first<00:03:30.200> wouldn't<00:03:30.640> be 00:03:31.630 --> 00:03:31.640 align:start position:0% so uh highest score first wouldn't be 00:03:31.640 --> 00:03:35.830 align:start position:0% so uh highest score first wouldn't be the<00:03:31.959> most<00:03:32.959> uh<00:03:33.640> successful<00:03:34.280> strategy<00:03:35.280> in<00:03:35.680> in 00:03:35.830 --> 00:03:35.840 align:start position:0% the most uh successful strategy in in 00:03:35.840 --> 00:03:37.350 align:start position:0% the most uh successful strategy in in some<00:03:36.120> cases<00:03:36.560> right<00:03:36.760> you<00:03:36.840> would<00:03:37.000> want<00:03:37.159> to<00:03:37.239> go 00:03:37.350 --> 00:03:37.360 align:start position:0% some cases right you would want to go 00:03:37.360 --> 00:03:38.710 align:start position:0% some cases right you would want to go for<00:03:37.519> the<00:03:37.599> average<00:03:37.920> score<00:03:38.200> because<00:03:38.400> you<00:03:38.519> know 00:03:38.710 --> 00:03:38.720 align:start position:0% for the average score because you know 00:03:38.720 --> 00:03:40.910 align:start position:0% for the average score because you know that<00:03:38.879> there<00:03:39.000> are<00:03:39.519> some<00:03:39.720> honest<00:03:40.120> nodes 00:03:40.910 --> 00:03:40.920 align:start position:0% that there are some honest nodes 00:03:40.920 --> 00:03:43.390 align:start position:0% that there are some honest nodes balancing<00:03:41.439> out<00:03:41.840> the<00:03:42.400> uh<00:03:42.560> malicious<00:03:43.040> nodes<00:03:43.280> in 00:03:43.390 --> 00:03:43.400 align:start position:0% balancing out the uh malicious nodes in 00:03:43.400 --> 00:03:45.190 align:start position:0% balancing out the uh malicious nodes in that<00:03:43.560> sub<00:03:43.799> tree<00:03:44.280> right<00:03:44.720> so<00:03:44.959> there<00:03:45.040> are 00:03:45.190 --> 00:03:45.200 align:start position:0% that sub tree right so there are 00:03:45.200 --> 00:03:46.830 align:start position:0% that sub tree right so there are different<00:03:45.560> sing<00:03:45.920> strategies<00:03:46.360> that<00:03:46.480> you<00:03:46.560> can 00:03:46.830 --> 00:03:46.840 align:start position:0% different sing strategies that you can 00:03:46.840 --> 00:03:49.509 align:start position:0% different sing strategies that you can apply<00:03:47.840> uh<00:03:48.239> and<00:03:48.439> for<00:03:48.599> the<00:03:48.720> rated<00:03:49.040> list<00:03:49.280> we<00:03:49.360> have 00:03:49.509 --> 00:03:49.519 align:start position:0% apply uh and for the rated list we have 00:03:49.519 --> 00:03:51.229 align:start position:0% apply uh and for the rated list we have defined<00:03:49.799> a<00:03:49.920> default<00:03:50.200> score<00:03:50.439> of<00:03:50.560> 1.0<00:03:51.040> which<00:03:51.159> is 00:03:51.229 --> 00:03:51.239 align:start position:0% defined a default score of 1.0 which is 00:03:51.239 --> 00:03:53.069 align:start position:0% defined a default score of 1.0 which is a<00:03:51.439> perfect<00:03:51.760> score<00:03:52.360> uh<00:03:52.480> which<00:03:52.599> is<00:03:52.720> slightly 00:03:53.069 --> 00:03:53.079 align:start position:0% a perfect score uh which is slightly 00:03:53.079 --> 00:03:55.630 align:start position:0% a perfect score uh which is slightly optimistic<00:03:53.519> but<00:03:53.640> it's<00:03:53.879> fine<00:03:54.519> because<00:03:55.159> uh<00:03:55.519> we 00:03:55.630 --> 00:03:55.640 align:start position:0% optimistic but it's fine because uh we 00:03:55.640 --> 00:03:59.429 align:start position:0% optimistic but it's fine because uh we do<00:03:55.799> scoring<00:03:56.159> on<00:03:56.280> a<00:03:56.480> per<00:03:56.840> slot<00:03:57.480> basis<00:03:58.480> so<00:03:59.000> uh<00:03:59.280> for 00:03:59.429 --> 00:03:59.439 align:start position:0% do scoring on a per slot basis so uh for 00:03:59.439 --> 00:04:03.229 align:start position:0% do scoring on a per slot basis so uh for every<00:03:59.879> slot<00:04:00.720> we<00:04:01.319> uh<00:04:01.439> get<00:04:01.640> all<00:04:01.799> the<00:04:01.959> peers<00:04:02.879> uh<00:04:03.040> of 00:04:03.229 --> 00:04:03.239 align:start position:0% every slot we uh get all the peers uh of 00:04:03.239 --> 00:04:05.789 align:start position:0% every slot we uh get all the peers uh of peers<00:04:03.599> and<00:04:03.720> build<00:04:03.959> the<00:04:04.079> tree<00:04:04.840> and<00:04:05.319> uh<00:04:05.439> we<00:04:05.560> start 00:04:05.789 --> 00:04:05.799 align:start position:0% peers and build the tree and uh we start 00:04:05.799 --> 00:04:08.229 align:start position:0% peers and build the tree and uh we start scoring<00:04:06.319> and<00:04:06.439> we<00:04:06.599> delete<00:04:06.920> the<00:04:07.040> scores<00:04:07.920> for<00:04:08.079> the 00:04:08.229 --> 00:04:08.239 align:start position:0% scoring and we delete the scores for the 00:04:08.239 --> 00:04:10.670 align:start position:0% scoring and we delete the scores for the next<00:04:08.439> slot<00:04:08.760> we<00:04:08.879> start<00:04:09.239> again<00:04:09.640> right<00:04:10.480> uh 00:04:10.670 --> 00:04:10.680 align:start position:0% next slot we start again right uh 00:04:10.680 --> 00:04:12.750 align:start position:0% next slot we start again right uh ideally<00:04:11.120> we<00:04:11.280> could<00:04:11.640> what<00:04:11.799> we<00:04:11.920> could<00:04:12.159> do<00:04:12.400> is<00:04:12.640> we 00:04:12.750 --> 00:04:12.760 align:start position:0% ideally we could what we could do is we 00:04:12.760 --> 00:04:14.309 align:start position:0% ideally we could what we could do is we could<00:04:12.959> propagate<00:04:13.480> these<00:04:13.640> scores<00:04:13.959> into<00:04:14.159> the 00:04:14.309 --> 00:04:14.319 align:start position:0% could propagate these scores into the 00:04:14.319 --> 00:04:17.749 align:start position:0% could propagate these scores into the gossip<00:04:14.640> score<00:04:15.159> uh<00:04:15.479> v1.2<00:04:16.479> to<00:04:16.759> persist<00:04:17.199> over 00:04:17.749 --> 00:04:17.759 align:start position:0% gossip score uh v1.2 to persist over 00:04:17.759 --> 00:04:20.469 align:start position:0% gossip score uh v1.2 to persist over slots<00:04:18.759> uh<00:04:18.880> so<00:04:19.079> we<00:04:19.199> could<00:04:19.400> have<00:04:19.680> that<00:04:19.840> kind<00:04:20.000> of<00:04:20.160> a 00:04:20.469 --> 00:04:20.479 align:start position:0% slots uh so we could have that kind of a 00:04:20.479 --> 00:04:22.950 align:start position:0% slots uh so we could have that kind of a scoring<00:04:20.840> mechanism<00:04:21.359> as<00:04:21.519> well<00:04:22.360> uh<00:04:22.560> why<00:04:22.720> do<00:04:22.840> we 00:04:22.950 --> 00:04:22.960 align:start position:0% scoring mechanism as well uh why do we 00:04:22.960 --> 00:04:25.189 align:start position:0% scoring mechanism as well uh why do we use<00:04:23.160> the<00:04:23.280> rated<00:04:23.600> list<00:04:24.080> right<00:04:24.520> uh<00:04:24.680> the<00:04:24.840> main 00:04:25.189 --> 00:04:25.199 align:start position:0% use the rated list right uh the main 00:04:25.199 --> 00:04:27.870 align:start position:0% use the rated list right uh the main attack<00:04:25.560> that<00:04:25.720> we<00:04:25.840> are<00:04:26.400> uh<00:04:26.880> going<00:04:27.199> towards<00:04:27.720> is 00:04:27.870 --> 00:04:27.880 align:start position:0% attack that we are uh going towards is 00:04:27.880 --> 00:04:30.950 align:start position:0% attack that we are uh going towards is to<00:04:28.759> defend<00:04:29.240> against<00:04:29.759> a<00:04:29.880> local<00:04:30.160> keyspace<00:04:30.880> uh 00:04:30.950 --> 00:04:30.960 align:start position:0% to defend against a local keyspace uh 00:04:30.960 --> 00:04:34.469 align:start position:0% to defend against a local keyspace uh flooding<00:04:31.400> attack<00:04:32.360> uh<00:04:32.479> through<00:04:32.680> Cil<00:04:33.120> notes<00:04:34.120> uh 00:04:34.469 --> 00:04:34.479 align:start position:0% flooding attack uh through Cil notes uh 00:04:34.479 --> 00:04:36.870 align:start position:0% flooding attack uh through Cil notes uh we<00:04:34.720> also<00:04:35.000> want<00:04:35.199> to<00:04:35.479> reduce<00:04:36.120> the<00:04:36.280> amount<00:04:36.600> of 00:04:36.870 --> 00:04:36.880 align:start position:0% we also want to reduce the amount of 00:04:36.880 --> 00:04:39.550 align:start position:0% we also want to reduce the amount of requests<00:04:37.360> that<00:04:37.479> we<00:04:37.680> sent<00:04:38.120> out<00:04:39.120> uh<00:04:39.280> for 00:04:39.550 --> 00:04:39.560 align:start position:0% requests that we sent out uh for 00:04:39.560 --> 00:04:41.350 align:start position:0% requests that we sent out uh for actually<00:04:39.880> completing<00:04:40.280> pure<00:04:40.560> sampling<00:04:41.199> which 00:04:41.350 --> 00:04:41.360 align:start position:0% actually completing pure sampling which 00:04:41.360 --> 00:04:44.790 align:start position:0% actually completing pure sampling which can<00:04:41.479> be<00:04:41.680> avoided<00:04:42.120> if<00:04:42.560> we<00:04:42.720> can<00:04:43.520> defunct<00:04:44.479> uh<00:04:44.600> an 00:04:44.790 --> 00:04:44.800 align:start position:0% can be avoided if we can defunct uh an 00:04:44.800 --> 00:04:48.790 align:start position:0% can be avoided if we can defunct uh an entire<00:04:45.120> subtree<00:04:45.720> of<00:04:45.880> malicious<00:04:46.600> nodes<00:04:47.720> um<00:04:48.720> I 00:04:48.790 --> 00:04:48.800 align:start position:0% entire subtree of malicious nodes um I 00:04:48.800 --> 00:04:51.230 align:start position:0% entire subtree of malicious nodes um I would<00:04:49.039> say<00:04:49.320> that<00:04:49.440> the<00:04:49.600> P<00:04:49.880> slot<00:04:50.160> Tempo<00:04:50.639> matches 00:04:51.230 --> 00:04:51.240 align:start position:0% would say that the P slot Tempo matches 00:04:51.240 --> 00:04:54.749 align:start position:0% would say that the P slot Tempo matches very<00:04:51.520> nicely<00:04:52.120> with<00:04:52.880> uh<00:04:53.880> with<00:04:54.240> uh<00:04:54.320> the<00:04:54.440> peer 00:04:54.749 --> 00:04:54.759 align:start position:0% very nicely with uh with uh the peer 00:04:54.759 --> 00:04:58.950 align:start position:0% very nicely with uh with uh the peer samp<00:04:55.199> sampling<00:04:56.000> requirements<00:04:57.000> uh<00:04:57.320> because<00:04:58.160> um 00:04:58.950 --> 00:04:58.960 align:start position:0% samp sampling requirements uh because um 00:04:58.960 --> 00:05:00.150 align:start position:0% samp sampling requirements uh because um you<00:04:59.120> can<00:04:59.360> have 00:05:00.150 --> 00:05:00.160 align:start position:0% you can have 00:05:00.160 --> 00:05:02.670 align:start position:0% you can have honest<00:05:00.800> uh<00:05:00.919> honest<00:05:01.240> acting<00:05:01.520> malicious<00:05:02.000> nodes 00:05:02.670 --> 00:05:02.680 align:start position:0% honest uh honest acting malicious nodes 00:05:02.680 --> 00:05:05.310 align:start position:0% honest uh honest acting malicious nodes which<00:05:02.840> is<00:05:03.039> default<00:05:03.479> on<00:05:03.720> one<00:05:04.199> particular<00:05:04.680> block 00:05:05.310 --> 00:05:05.320 align:start position:0% which is default on one particular block 00:05:05.320 --> 00:05:07.830 align:start position:0% which is default on one particular block right<00:05:05.560> so<00:05:06.000> the<00:05:06.120> entire<00:05:06.479> objective<00:05:07.000> is<00:05:07.680> to 00:05:07.830 --> 00:05:07.840 align:start position:0% right so the entire objective is to 00:05:07.840 --> 00:05:09.310 align:start position:0% right so the entire objective is to complete<00:05:08.160> sampling<00:05:08.560> for<00:05:08.759> that<00:05:08.960> particular 00:05:09.310 --> 00:05:09.320 align:start position:0% complete sampling for that particular 00:05:09.320 --> 00:05:11.029 align:start position:0% complete sampling for that particular slot<00:05:09.600> so<00:05:09.720> up<00:05:09.880> per<00:05:10.000> slot<00:05:10.240> sampling<00:05:10.639> sets<00:05:10.919> the 00:05:11.029 --> 00:05:11.039 align:start position:0% slot so up per slot sampling sets the 00:05:11.039 --> 00:05:14.430 align:start position:0% slot so up per slot sampling sets the tempo<00:05:11.520> right<00:05:12.520> uh<00:05:13.000> and<00:05:13.240> lastly<00:05:13.880> it<00:05:14.039> removes 00:05:14.430 --> 00:05:14.440 align:start position:0% tempo right uh and lastly it removes 00:05:14.440 --> 00:05:16.790 align:start position:0% tempo right uh and lastly it removes dead<00:05:14.680> sub<00:05:14.919> trees<00:05:15.320> so<00:05:15.960> if<00:05:16.039> there<00:05:16.160> are<00:05:16.360> any 00:05:16.790 --> 00:05:16.800 align:start position:0% dead sub trees so if there are any 00:05:16.800 --> 00:05:18.469 align:start position:0% dead sub trees so if there are any inactive<00:05:17.240> nodes<00:05:17.759> they<00:05:17.840> would<00:05:18.039> already<00:05:18.319> be 00:05:18.469 --> 00:05:18.479 align:start position:0% inactive nodes they would already be 00:05:18.479 --> 00:05:20.950 align:start position:0% inactive nodes they would already be removed<00:05:19.039> and<00:05:19.720> you<00:05:19.880> could<00:05:20.280> you<00:05:20.440> could<00:05:20.639> assume 00:05:20.950 --> 00:05:20.960 align:start position:0% removed and you could you could assume 00:05:20.960 --> 00:05:22.870 align:start position:0% removed and you could you could assume that<00:05:21.160> we<00:05:21.280> reach<00:05:21.520> a<00:05:21.680> global<00:05:22.000> stability<00:05:22.520> Point 00:05:22.870 --> 00:05:22.880 align:start position:0% that we reach a global stability Point 00:05:22.880 --> 00:05:24.430 align:start position:0% that we reach a global stability Point like<00:05:23.080> where<00:05:23.240> everyone<00:05:23.680> removes<00:05:23.960> that<00:05:24.080> defunct 00:05:24.430 --> 00:05:24.440 align:start position:0% like where everyone removes that defunct 00:05:24.440 --> 00:05:25.710 align:start position:0% like where everyone removes that defunct sub 00:05:25.710 --> 00:05:25.720 align:start position:0% sub 00:05:25.720 --> 00:05:29.029 align:start position:0% sub trees<00:05:26.720> uh<00:05:27.440> so<00:05:27.800> then<00:05:28.000> coming<00:05:28.639> uh<00:05:28.759> to<00:05:28.880> the 00:05:29.029 --> 00:05:29.039 align:start position:0% trees uh so then coming uh to the 00:05:29.039 --> 00:05:30.950 align:start position:0% trees uh so then coming uh to the simulator<00:05:29.479> and<00:05:29.720> the<00:05:29.800> simulator's<00:05:30.360> design<00:05:30.800> we 00:05:30.950 --> 00:05:30.960 align:start position:0% simulator and the simulator's design we 00:05:30.960 --> 00:05:34.590 align:start position:0% simulator and the simulator's design we had<00:05:31.080> to<00:05:31.199> play<00:05:31.400> around<00:05:31.680> a<00:05:31.800> lot<00:05:32.720> uh<00:05:32.919> because<00:05:33.680> for 00:05:34.590 --> 00:05:34.600 align:start position:0% had to play around a lot uh because for 00:05:34.600 --> 00:05:38.110 align:start position:0% had to play around a lot uh because for graphs<00:05:35.600> uh<00:05:36.280> they're<00:05:36.600> not<00:05:37.319> uh<00:05:37.759> there<00:05:37.880> are 00:05:38.110 --> 00:05:38.120 align:start position:0% graphs uh they're not uh there are 00:05:38.120 --> 00:05:39.909 align:start position:0% graphs uh they're not uh there are libraries<00:05:38.600> out<00:05:38.840> there<00:05:39.080> but<00:05:39.440> the<00:05:39.560> ones<00:05:39.800> that 00:05:39.909 --> 00:05:39.919 align:start position:0% libraries out there but the ones that 00:05:39.919 --> 00:05:42.749 align:start position:0% libraries out there but the ones that are<00:05:40.080> usually<00:05:40.479> used<00:05:41.160> uh<00:05:41.280> are<00:05:41.479> not<00:05:41.840> that<00:05:42.479> uh 00:05:42.749 --> 00:05:42.759 align:start position:0% are usually used uh are not that uh 00:05:42.759 --> 00:05:44.830 align:start position:0% are usually used uh are not that uh optimized<00:05:43.560> so<00:05:43.720> we<00:05:43.840> started<00:05:44.199> with<00:05:44.319> the<00:05:44.400> network 00:05:44.830 --> 00:05:44.840 align:start position:0% optimized so we started with the network 00:05:44.840 --> 00:05:48.909 align:start position:0% optimized so we started with the network X<00:05:45.360> Library<00:05:46.360> but<00:05:47.120> uh<00:05:47.680> it<00:05:47.840> was<00:05:48.160> really<00:05:48.400> slow<00:05:48.720> at 00:05:48.909 --> 00:05:48.919 align:start position:0% X Library but uh it was really slow at 00:05:48.919 --> 00:05:50.670 align:start position:0% X Library but uh it was really slow at generating<00:05:49.360> a<00:05:49.520> random<00:05:49.880> graph<00:05:50.120> with<00:05:50.240> a<00:05:50.400> high 00:05:50.670 --> 00:05:50.680 align:start position:0% generating a random graph with a high 00:05:50.680 --> 00:05:53.590 align:start position:0% generating a random graph with a high degree<00:05:51.680> uh<00:05:51.840> at<00:05:52.000> number<00:05:52.240> of<00:05:52.360> nodes<00:05:52.680> at<00:05:53.039> you<00:05:53.160> know 00:05:53.590 --> 00:05:53.600 align:start position:0% degree uh at number of nodes at you know 00:05:53.600 --> 00:05:55.990 align:start position:0% degree uh at number of nodes at you know 10,000<00:05:54.600> right<00:05:54.800> now<00:05:54.960> for<00:05:55.120> all<00:05:55.280> our<00:05:55.479> simulations 00:05:55.990 --> 00:05:56.000 align:start position:0% 10,000 right now for all our simulations 00:05:56.000 --> 00:05:58.230 align:start position:0% 10,000 right now for all our simulations we<00:05:56.160> use<00:05:56.400> the<00:05:56.520> degree<00:05:56.880> at<00:05:57.360> 50<00:05:57.800> and<00:05:57.960> the<00:05:58.039> number 00:05:58.230 --> 00:05:58.240 align:start position:0% we use the degree at 50 and the number 00:05:58.240 --> 00:06:00.830 align:start position:0% we use the degree at 50 and the number of<00:05:58.360> nodes<00:05:58.639> at<00:05:58.759> 10,000<00:05:59.759> which<00:05:59.919> is<00:06:00.039> like<00:06:00.199> a<00:06:00.600> good 00:06:00.830 --> 00:06:00.840 align:start position:0% of nodes at 10,000 which is like a good 00:06:00.840 --> 00:06:03.830 align:start position:0% of nodes at 10,000 which is like a good assumption<00:06:01.280> for<00:06:01.479> a<00:06:01.639> p<00:06:02.120> sampling<00:06:02.880> uh<00:06:03.120> Network 00:06:03.830 --> 00:06:03.840 align:start position:0% assumption for a p sampling uh Network 00:06:03.840 --> 00:06:07.350 align:start position:0% assumption for a p sampling uh Network right<00:06:04.800> um<00:06:05.280> and<00:06:05.400> we<00:06:05.520> use<00:06:06.199> R<00:06:06.560> Rush<00:06:06.880> workor<00:06:07.080> X 00:06:07.350 --> 00:06:07.360 align:start position:0% right um and we use R Rush workor X 00:06:07.360 --> 00:06:09.870 align:start position:0% right um and we use R Rush workor X because<00:06:08.160> uh<00:06:08.319> as<00:06:08.400> you<00:06:08.520> can<00:06:08.639> see<00:06:08.840> the<00:06:09.000> benchmarks 00:06:09.870 --> 00:06:09.880 align:start position:0% because uh as you can see the benchmarks 00:06:09.880 --> 00:06:12.309 align:start position:0% because uh as you can see the benchmarks for<00:06:10.319> uh<00:06:10.440> graph<00:06:10.759> generation<00:06:11.199> is<00:06:11.479> just<00:06:12.000> uh<00:06:12.120> the 00:06:12.309 --> 00:06:12.319 align:start position:0% for uh graph generation is just uh the 00:06:12.319 --> 00:06:14.350 align:start position:0% for uh graph generation is just uh the best<00:06:13.160> for<00:06:13.360> all<00:06:13.520> the<00:06:13.639> libraries<00:06:14.080> that<00:06:14.199> are 00:06:14.350 --> 00:06:14.360 align:start position:0% best for all the libraries that are 00:06:14.360 --> 00:06:16.790 align:start position:0% best for all the libraries that are there<00:06:14.560> out<00:06:14.800> there<00:06:15.039> right<00:06:15.199> now<00:06:16.160> uh<00:06:16.400> and<00:06:16.560> for 00:06:16.790 --> 00:06:16.800 align:start position:0% there out there right now uh and for 00:06:16.800 --> 00:06:18.710 align:start position:0% there out there right now uh and for quering<00:06:17.199> strategy<00:06:17.680> we<00:06:17.840> used<00:06:18.280> all<00:06:18.479> four 00:06:18.710 --> 00:06:18.720 align:start position:0% quering strategy we used all four 00:06:18.720 --> 00:06:21.230 align:start position:0% quering strategy we used all four quering<00:06:19.080> strategies<00:06:19.680> highest<00:06:20.080> average<00:06:21.080> uh 00:06:21.230 --> 00:06:21.240 align:start position:0% quering strategies highest average uh 00:06:21.240 --> 00:06:24.510 align:start position:0% quering strategies highest average uh lowest<00:06:21.759> and<00:06:22.240> actually<00:06:22.479> random<00:06:22.919> as<00:06:23.120> well<00:06:24.080> uh 00:06:24.510 --> 00:06:24.520 align:start position:0% lowest and actually random as well uh 00:06:24.520 --> 00:06:27.710 align:start position:0% lowest and actually random as well uh but<00:06:25.520> we<00:06:25.880> also<00:06:26.280> ref<00:06:26.560> filter<00:06:27.000> nodes<00:06:27.440> with<00:06:27.560> a 00:06:27.710 --> 00:06:27.720 align:start position:0% but we also ref filter nodes with a 00:06:27.720 --> 00:06:29.189 align:start position:0% but we also ref filter nodes with a different<00:06:28.000> threshold<00:06:28.680> if<00:06:28.800> we<00:06:28.919> cannot 00:06:29.189 --> 00:06:29.199 align:start position:0% different threshold if we cannot 00:06:29.199 --> 00:06:31.790 align:start position:0% different threshold if we cannot complete<00:06:29.960> sampling<00:06:30.960> uh<00:06:31.280> because<00:06:31.560> our 00:06:31.790 --> 00:06:31.800 align:start position:0% complete sampling uh because our 00:06:31.800 --> 00:06:33.390 align:start position:0% complete sampling uh because our objective<00:06:32.160> is<00:06:32.280> to<00:06:32.400> complete<00:06:32.759> sampling<00:06:33.199> at<00:06:33.319> the 00:06:33.390 --> 00:06:33.400 align:start position:0% objective is to complete sampling at the 00:06:33.400 --> 00:06:37.029 align:start position:0% objective is to complete sampling at the end<00:06:33.560> of<00:06:33.680> the<00:06:33.840> day<00:06:34.440> right<00:06:34.960> so<00:06:35.599> um<00:06:36.560> uh<00:06:36.800> coming 00:06:37.029 --> 00:06:37.039 align:start position:0% end of the day right so um uh coming 00:06:37.039 --> 00:06:38.629 align:start position:0% end of the day right so um uh coming into<00:06:37.240> the<00:06:37.400> architecture<00:06:37.919> of<00:06:38.039> the<00:06:38.120> simulator<00:06:38.560> a 00:06:38.629 --> 00:06:38.639 align:start position:0% into the architecture of the simulator a 00:06:38.639 --> 00:06:41.070 align:start position:0% into the architecture of the simulator a little<00:06:38.840> bit<00:06:39.520> uh<00:06:39.800> we<00:06:39.960> wanted<00:06:40.280> to<00:06:40.400> modularize 00:06:41.070 --> 00:06:41.080 align:start position:0% little bit uh we wanted to modularize 00:06:41.080 --> 00:06:44.430 align:start position:0% little bit uh we wanted to modularize everything<00:06:41.479> so<00:06:41.680> that<00:06:42.440> this<00:06:43.039> simulator<00:06:44.039> can 00:06:44.430 --> 00:06:44.440 align:start position:0% everything so that this simulator can 00:06:44.440 --> 00:06:46.629 align:start position:0% everything so that this simulator can practically<00:06:44.960> be<00:06:45.160> used<00:06:45.560> for<00:06:45.880> other<00:06:46.160> kinds<00:06:46.440> of 00:06:46.629 --> 00:06:46.639 align:start position:0% practically be used for other kinds of 00:06:46.639 --> 00:06:48.670 align:start position:0% practically be used for other kinds of implementations<00:06:47.440> like<00:06:47.599> the<00:06:47.720> rated<00:06:48.080> list<00:06:48.520> in 00:06:48.670 --> 00:06:48.680 align:start position:0% implementations like the rated list in 00:06:48.680 --> 00:06:51.830 align:start position:0% implementations like the rated list in the<00:06:48.800> future<00:06:49.639> so<00:06:50.360> uh<00:06:50.599> having<00:06:50.840> it<00:06:51.000> modularized 00:06:51.830 --> 00:06:51.840 align:start position:0% the future so uh having it modularized 00:06:51.840 --> 00:06:54.270 align:start position:0% the future so uh having it modularized make<00:06:52.639> uh<00:06:52.800> gives<00:06:53.039> you<00:06:53.560> attack<00:06:53.800> framework 00:06:54.270 --> 00:06:54.280 align:start position:0% make uh gives you attack framework 00:06:54.280 --> 00:06:55.830 align:start position:0% make uh gives you attack framework because<00:06:54.479> we<00:06:54.560> have<00:06:54.680> written<00:06:54.919> a<00:06:55.039> lot<00:06:55.199> of<00:06:55.400> attacks 00:06:55.830 --> 00:06:55.840 align:start position:0% because we have written a lot of attacks 00:06:55.840 --> 00:06:58.189 align:start position:0% because we have written a lot of attacks right<00:06:56.000> now<00:06:56.680> uh<00:06:56.759> so<00:06:56.919> you<00:06:57.039> could<00:06:57.280> test<00:06:57.759> against 00:06:58.189 --> 00:06:58.199 align:start position:0% right now uh so you could test against 00:06:58.199 --> 00:07:02.390 align:start position:0% right now uh so you could test against these<00:06:58.960> um<00:06:59.599> in<00:06:59.680> the<00:06:59.879> future<00:07:00.879> uh<00:07:01.080> so<00:07:01.800> yeah 00:07:02.390 --> 00:07:02.400 align:start position:0% these um in the future uh so yeah 00:07:02.400 --> 00:07:04.790 align:start position:0% these um in the future uh so yeah everything's<00:07:02.919> modularized<00:07:03.840> we<00:07:04.080> spec<00:07:04.440> convert 00:07:04.790 --> 00:07:04.800 align:start position:0% everything's modularized we spec convert 00:07:04.800 --> 00:07:07.749 align:start position:0% everything's modularized we spec convert the<00:07:04.919> rated<00:07:05.319> list<00:07:05.840> uh<00:07:06.039> spec<00:07:06.479> into<00:07:06.720> a<00:07:06.840> notep 00:07:07.749 --> 00:07:07.759 align:start position:0% the rated list uh spec into a notep 00:07:07.759 --> 00:07:10.029 align:start position:0% the rated list uh spec into a notep implementation<00:07:08.759> for<00:07:08.960> all<00:07:09.080> the<00:07:09.240> simulations 00:07:10.029 --> 00:07:10.039 align:start position:0% implementation for all the simulations 00:07:10.039 --> 00:07:12.550 align:start position:0% implementation for all the simulations I'll<00:07:10.240> give<00:07:10.360> it<00:07:10.520> to 00:07:12.550 --> 00:07:12.560 align:start position:0% I'll give it to 00:07:12.560 --> 00:07:19.029 align:start position:0% I'll give it to open<00:07:13.560> so<00:07:14.240> um<00:07:15.080> yeah<00:07:16.080> that's<00:07:16.360> the 00:07:19.029 --> 00:07:19.039 align:start position:0% 00:07:19.039 --> 00:07:22.270 align:start position:0% simulator<00:07:20.039> let<00:07:20.199> me<00:07:20.479> just 00:07:22.270 --> 00:07:22.280 align:start position:0% simulator let me just 00:07:22.280 --> 00:07:28.510 align:start position:0% simulator let me just quickly 00:07:28.510 --> 00:07:28.520 align:start position:0% 00:07:28.520 --> 00:07:42.830 align:start position:0% set<00:07:29.520> I<00:07:29.680> think<00:07:29.800> it's 00:07:42.830 --> 00:07:42.840 align:start position:0% 00:07:42.840 --> 00:07:52.710 align:start position:0% okay<00:07:43.840> okay 00:07:52.710 --> 00:07:52.720 align:start position:0% 00:07:52.720 --> 00:07:59.550 align:start position:0% um<00:07:53.720> um 00:07:59.550 --> 00:07:59.560 align:start position:0% 00:07:59.560 --> 00:08:02.149 align:start position:0% start<00:07:59.759> explaining<00:08:00.120> the<00:08:00.280> attack<00:08:00.960> um<00:08:01.599> I<00:08:01.680> mean 00:08:02.149 --> 00:08:02.159 align:start position:0% start explaining the attack um I mean 00:08:02.159 --> 00:08:05.230 align:start position:0% start explaining the attack um I mean just<00:08:02.759> since<00:08:03.199> the<00:08:03.440> slides<00:08:03.800> are<00:08:04.039> off<00:08:04.440> let's<00:08:05.000> uh 00:08:05.230 --> 00:08:05.240 align:start position:0% just since the slides are off let's uh 00:08:05.240 --> 00:08:09.309 align:start position:0% just since the slides are off let's uh turn<00:08:05.520> the<00:08:05.720> attacks<00:08:06.039> on<00:08:06.319> I<00:08:07.000> guess<00:08:08.000> so<00:08:08.560> uh<00:08:08.800> yeah 00:08:09.309 --> 00:08:09.319 align:start position:0% turn the attacks on I guess so uh yeah 00:08:09.319 --> 00:08:12.230 align:start position:0% turn the attacks on I guess so uh yeah I'll<00:08:10.159> uh<00:08:10.319> quickly<00:08:10.599> brief<00:08:11.039> you<00:08:11.240> guys<00:08:11.440> about<00:08:12.120> uh 00:08:12.230 --> 00:08:12.240 align:start position:0% I'll uh quickly brief you guys about uh 00:08:12.240 --> 00:08:15.950 align:start position:0% I'll uh quickly brief you guys about uh the<00:08:12.360> sort<00:08:12.599> of<00:08:12.759> attacks<00:08:13.159> we<00:08:13.319> try<00:08:13.680> to<00:08:14.680> um<00:08:15.400> uh<00:08:15.599> use 00:08:15.950 --> 00:08:15.960 align:start position:0% the sort of attacks we try to um uh use 00:08:15.960 --> 00:08:17.510 align:start position:0% the sort of attacks we try to um uh use against<00:08:16.360> our<00:08:16.840> uh 00:08:17.510 --> 00:08:17.520 align:start position:0% against our uh 00:08:17.520 --> 00:08:19.990 align:start position:0% against our uh construction 00:08:19.990 --> 00:08:20.000 align:start position:0% construction 00:08:20.000 --> 00:08:24.070 align:start position:0% construction uh<00:08:21.000> okay<00:08:21.400> uh<00:08:21.599> yeah<00:08:21.840> so<00:08:22.840> uh<00:08:23.080> basically<00:08:23.639> what<00:08:23.879> we 00:08:24.070 --> 00:08:24.080 align:start position:0% uh okay uh yeah so uh basically what we 00:08:24.080 --> 00:08:27.510 align:start position:0% uh okay uh yeah so uh basically what we Tred<00:08:24.319> to<00:08:24.520> do<00:08:24.879> is<00:08:25.440> uh<00:08:25.639> poison<00:08:26.080> the<00:08:26.360> entire<00:08:27.360> uh 00:08:27.510 --> 00:08:27.520 align:start position:0% Tred to do is uh poison the entire uh 00:08:27.520 --> 00:08:31.230 align:start position:0% Tred to do is uh poison the entire uh network<00:08:28.120> with<00:08:28.520> a<00:08:28.680> lot<00:08:28.879> of<00:08:29.080> Random<00:08:29.800> picked<00:08:30.800> uh 00:08:31.230 --> 00:08:31.240 align:start position:0% network with a lot of Random picked uh 00:08:31.240 --> 00:08:35.350 align:start position:0% network with a lot of Random picked uh nodes<00:08:32.240> uh<00:08:32.519> and<00:08:32.719> mark<00:08:33.120> them<00:08:33.279> as<00:08:33.800> malicious<00:08:34.800> uh 00:08:35.350 --> 00:08:35.360 align:start position:0% nodes uh and mark them as malicious uh 00:08:35.360 --> 00:08:37.829 align:start position:0% nodes uh and mark them as malicious uh the<00:08:35.479> only<00:08:35.760> problem<00:08:36.039> with<00:08:36.360> that<00:08:36.680> is<00:08:36.959> that<00:08:37.719> uh 00:08:37.829 --> 00:08:37.839 align:start position:0% the only problem with that is that uh 00:08:37.839 --> 00:08:40.469 align:start position:0% the only problem with that is that uh the<00:08:38.120> conditional<00:08:38.519> probability<00:08:39.240> of<00:08:39.479> finding 00:08:40.469 --> 00:08:40.479 align:start position:0% the conditional probability of finding 00:08:40.479 --> 00:08:43.310 align:start position:0% the conditional probability of finding uh<00:08:40.640> the<00:08:40.800> parent<00:08:41.200> being<00:08:41.959> malicious<00:08:42.959> uh<00:08:43.120> with 00:08:43.310 --> 00:08:43.320 align:start position:0% uh the parent being malicious uh with 00:08:43.320 --> 00:08:46.750 align:start position:0% uh the parent being malicious uh with the<00:08:43.959> uh<00:08:44.120> children<00:08:44.959> is<00:08:45.200> almost<00:08:45.760> same<00:08:46.200> as<00:08:46.519> just 00:08:46.750 --> 00:08:46.760 align:start position:0% the uh children is almost same as just 00:08:46.760 --> 00:08:49.990 align:start position:0% the uh children is almost same as just randomly<00:08:47.640> picking<00:08:48.080> any<00:08:48.760> uh<00:08:48.880> malicious<00:08:49.440> node 00:08:49.990 --> 00:08:50.000 align:start position:0% randomly picking any uh malicious node 00:08:50.000 --> 00:08:52.750 align:start position:0% randomly picking any uh malicious node from<00:08:50.160> the<00:08:50.320> network<00:08:50.760> so<00:08:51.600> uh<00:08:52.040> just<00:08:52.279> randomly 00:08:52.750 --> 00:08:52.760 align:start position:0% from the network so uh just randomly 00:08:52.760 --> 00:08:56.509 align:start position:0% from the network so uh just randomly marking<00:08:53.240> any<00:08:53.959> uh<00:08:54.440> node<00:08:55.440> uh<00:08:55.920> doesn't<00:08:56.279> make 00:08:56.509 --> 00:08:56.519 align:start position:0% marking any uh node uh doesn't make 00:08:56.519 --> 00:08:58.990 align:start position:0% marking any uh node uh doesn't make sense<00:08:56.920> and<00:08:57.440> like<00:08:57.839> the<00:08:58.040> there's<00:08:58.399> not<00:08:58.600> a<00:08:58.720> lot<00:08:58.880> of 00:08:58.990 --> 00:08:59.000 align:start position:0% sense and like the there's not a lot of 00:08:59.000 --> 00:09:01.829 align:start position:0% sense and like the there's not a lot of difference<00:08:59.519> because<00:08:59.760> of<00:08:59.880> the<00:09:00.000> rated<00:09:00.440> list<00:09:01.440> uh 00:09:01.829 --> 00:09:01.839 align:start position:0% difference because of the rated list uh 00:09:01.839 --> 00:09:05.150 align:start position:0% difference because of the rated list uh although<00:09:02.839> uh<00:09:03.000> the<00:09:03.279> only<00:09:03.640> attack<00:09:03.959> Vector<00:09:04.480> that 00:09:05.150 --> 00:09:05.160 align:start position:0% although uh the only attack Vector that 00:09:05.160 --> 00:09:09.910 align:start position:0% although uh the only attack Vector that uh<00:09:05.519> that<00:09:05.800> might<00:09:06.120> cause<00:09:06.600> an<00:09:06.959> issue<00:09:07.959> uh<00:09:08.560> would<00:09:08.920> be 00:09:09.910 --> 00:09:09.920 align:start position:0% uh that might cause an issue uh would be 00:09:09.920 --> 00:09:13.110 align:start position:0% uh that might cause an issue uh would be uh<00:09:10.640> basically<00:09:11.079> trying<00:09:11.360> to<00:09:12.079> attack<00:09:12.519> on<00:09:12.760> one 00:09:13.110 --> 00:09:13.120 align:start position:0% uh basically trying to attack on one 00:09:13.120 --> 00:09:15.949 align:start position:0% uh basically trying to attack on one local<00:09:13.480> key<00:09:13.800> space<00:09:14.640> and<00:09:15.200> uh<00:09:15.440> that's<00:09:15.640> what<00:09:15.800> we 00:09:15.949 --> 00:09:15.959 align:start position:0% local key space and uh that's what we 00:09:15.959 --> 00:09:16.829 align:start position:0% local key space and uh that's what we tried<00:09:16.200> to 00:09:16.829 --> 00:09:16.839 align:start position:0% tried to 00:09:16.839 --> 00:09:22.030 align:start position:0% tried to simulate<00:09:18.279> uh<00:09:19.279> so<00:09:20.120> yeah<00:09:21.120> basically<00:09:21.519> trying<00:09:21.800> to 00:09:22.030 --> 00:09:22.040 align:start position:0% simulate uh so yeah basically trying to 00:09:22.040 --> 00:09:25.829 align:start position:0% simulate uh so yeah basically trying to Mark<00:09:22.360> and<00:09:22.600> anise<00:09:23.519> subre<00:09:24.519> uh<00:09:24.760> to<00:09:25.279> a<00:09:25.440> malicious 00:09:25.829 --> 00:09:25.839 align:start position:0% Mark and anise subre uh to a malicious 00:09:25.839 --> 00:09:29.350 align:start position:0% Mark and anise subre uh to a malicious node<00:09:26.360> which<00:09:26.880> uh<00:09:27.120> then<00:09:28.000> gave<00:09:28.240> us<00:09:28.440> the<00:09:28.959> uh<00:09:29.120> result 00:09:29.350 --> 00:09:29.360 align:start position:0% node which uh then gave us the uh result 00:09:29.360 --> 00:09:32.030 align:start position:0% node which uh then gave us the uh result results<00:09:29.720> where<00:09:29.959> we<00:09:30.079> were<00:09:30.560> getting<00:09:31.560> uh<00:09:31.760> far 00:09:32.030 --> 00:09:32.040 align:start position:0% results where we were getting uh far 00:09:32.040 --> 00:09:34.630 align:start position:0% results where we were getting uh far more<00:09:32.360> better<00:09:32.680> results<00:09:33.079> in<00:09:33.200> terms<00:09:33.560> of<00:09:34.079> uh<00:09:34.200> rated 00:09:34.630 --> 00:09:34.640 align:start position:0% more better results in terms of uh rated 00:09:34.640 --> 00:09:37.509 align:start position:0% more better results in terms of uh rated list<00:09:35.480> as<00:09:35.640> compared<00:09:36.000> to<00:09:36.320> just 00:09:37.509 --> 00:09:37.519 align:start position:0% list as compared to just 00:09:37.519 --> 00:09:45.790 align:start position:0% list as compared to just randomly<00:09:38.519> yeah<00:09:38.959> there<00:09:39.079> you 00:09:45.790 --> 00:09:45.800 align:start position:0% 00:09:45.800 --> 00:09:50.030 align:start position:0% go<00:09:46.800> okay<00:09:47.320> uh<00:09:47.519> so<00:09:47.880> yeah<00:09:48.360> uh<00:09:48.880> there's<00:09:49.480> a<00:09:49.680> graph 00:09:50.030 --> 00:09:50.040 align:start position:0% go okay uh so yeah uh there's a graph 00:09:50.040 --> 00:09:54.230 align:start position:0% go okay uh so yeah uh there's a graph plot<00:09:50.519> where<00:09:50.800> we<00:09:51.640> uh<00:09:51.800> get<00:09:51.959> to<00:09:52.120> see<00:09:52.680> where<00:09:53.680> uh<00:09:54.079> the 00:09:54.230 --> 00:09:54.240 align:start position:0% plot where we uh get to see where uh the 00:09:54.240 --> 00:09:57.069 align:start position:0% plot where we uh get to see where uh the rated<00:09:54.600> list<00:09:55.160> performs<00:09:56.160> much<00:09:56.519> better<00:09:56.839> than 00:09:57.069 --> 00:09:57.079 align:start position:0% rated list performs much better than 00:09:57.079 --> 00:09:58.790 align:start position:0% rated list performs much better than just<00:09:57.320> randomly<00:09:57.839> picking<00:09:58.120> out<00:09:58.320> notes<00:09:58.640> and 00:09:58.790 --> 00:09:58.800 align:start position:0% just randomly picking out notes and 00:09:58.800 --> 00:10:02.230 align:start position:0% just randomly picking out notes and trying<00:09:59.000> to<00:09:59.519> example<00:10:00.519> uh<00:10:01.240> it's<00:10:01.440> in<00:10:01.600> terms<00:10:01.959> of 00:10:02.230 --> 00:10:02.240 align:start position:0% trying to example uh it's in terms of 00:10:02.240 --> 00:10:05.550 align:start position:0% trying to example uh it's in terms of the<00:10:02.360> number<00:10:02.600> of<00:10:02.839> requests<00:10:03.320> that<00:10:03.519> go<00:10:03.959> out 00:10:05.550 --> 00:10:05.560 align:start position:0% the number of requests that go out 00:10:05.560 --> 00:10:12.069 align:start position:0% the number of requests that go out uh<00:10:06.560> the<00:10:07.399> like<00:10:07.560> for<00:10:07.880> for<00:10:08.519> uh<00:10:08.760> 90%<00:10:09.560> of<00:10:10.279> 70%<00:10:11.200> of<00:10:11.760> uh 00:10:12.069 --> 00:10:12.079 align:start position:0% uh the like for for uh 90% of 70% of uh 00:10:12.079 --> 00:10:16.509 align:start position:0% uh the like for for uh 90% of 70% of uh civil<00:10:13.079> uh<00:10:13.240> nodes<00:10:14.240> gives<00:10:14.600> us<00:10:15.440> almost<00:10:15.959> double 00:10:16.509 --> 00:10:16.519 align:start position:0% civil uh nodes gives us almost double 00:10:16.519 --> 00:10:20.030 align:start position:0% civil uh nodes gives us almost double number<00:10:16.839> of<00:10:17.480> uh<00:10:18.200> requests<00:10:19.200> when<00:10:19.480> doing<00:10:19.800> a 00:10:20.030 --> 00:10:20.040 align:start position:0% number of uh requests when doing a 00:10:20.040 --> 00:10:23.030 align:start position:0% number of uh requests when doing a random<00:10:20.399> sampling<00:10:21.360> but<00:10:21.519> in<00:10:21.720> case<00:10:21.920> of<00:10:22.440> rated<00:10:22.800> L 00:10:23.030 --> 00:10:23.040 align:start position:0% random sampling but in case of rated L 00:10:23.040 --> 00:10:28.030 align:start position:0% random sampling but in case of rated L it's<00:10:23.200> like<00:10:23.839> uh<00:10:25.240> um<00:10:26.240> the<00:10:26.360> lower<00:10:26.640> number 00:10:28.030 --> 00:10:28.040 align:start position:0% it's like uh um the lower number 00:10:28.040 --> 00:10:31.030 align:start position:0% it's like uh um the lower number so<00:10:29.040> is<00:10:29.160> it 00:10:31.030 --> 00:10:31.040 align:start position:0% so is it 00:10:31.040 --> 00:10:34.350 align:start position:0% so is it okay<00:10:32.040> okay<00:10:32.279> got 00:10:34.350 --> 00:10:34.360 align:start position:0% okay okay got 00:10:34.360 --> 00:10:45.430 align:start position:0% okay okay got it<00:10:35.360> should<00:10:35.639> I 00:10:45.430 --> 00:10:45.440 align:start position:0% 00:10:45.440 --> 00:10:49.150 align:start position:0% okay<00:10:46.440> yeah<00:10:46.600> so<00:10:46.839> there<00:10:46.920> are<00:10:47.440> graphs<00:10:48.440> uh<00:10:48.680> showing 00:10:49.150 --> 00:10:49.160 align:start position:0% okay yeah so there are graphs uh showing 00:10:49.160 --> 00:10:51.949 align:start position:0% okay yeah so there are graphs uh showing those<00:10:49.440> results<00:10:49.959> and<00:10:50.600> uh<00:10:51.320> that<00:10:51.519> there's<00:10:51.720> some 00:10:51.949 --> 00:10:51.959 align:start position:0% those results and uh that there's some 00:10:51.959 --> 00:10:54.870 align:start position:0% those results and uh that there's some future<00:10:52.360> work<00:10:52.600> that<00:10:52.760> needs<00:10:52.959> to<00:10:53.079> be<00:10:53.480> done<00:10:54.480> uh<00:10:54.760> we 00:10:54.870 --> 00:10:54.880 align:start position:0% future work that needs to be done uh we 00:10:54.880 --> 00:10:56.670 align:start position:0% future work that needs to be done uh we are<00:10:55.000> currently<00:10:55.399> researching<00:10:55.880> on<00:10:56.120> this<00:10:56.360> new 00:10:56.670 --> 00:10:56.680 align:start position:0% are currently researching on this new 00:10:56.680 --> 00:11:01.790 align:start position:0% are currently researching on this new topic<00:10:57.519> uh<00:10:57.720> of<00:10:58.000> trying<00:10:58.320> to<00:10:59.320> you<00:10:59.680> uh<00:10:59.880> score<00:11:00.880> the 00:11:01.790 --> 00:11:01.800 align:start position:0% topic uh of trying to you uh score the 00:11:01.800 --> 00:11:04.470 align:start position:0% topic uh of trying to you uh score the parents<00:11:02.800> uh<00:11:03.160> in<00:11:03.279> a<00:11:03.440> probabilistic<00:11:04.120> way<00:11:04.320> right 00:11:04.470 --> 00:11:04.480 align:start position:0% parents uh in a probabilistic way right 00:11:04.480 --> 00:11:08.069 align:start position:0% parents uh in a probabilistic way right now<00:11:04.639> it's<00:11:04.839> just<00:11:05.040> averaging<00:11:05.680> out<00:11:06.680> uh<00:11:07.120> so<00:11:07.839> uh<00:11:07.959> we 00:11:08.069 --> 00:11:08.079 align:start position:0% now it's just averaging out uh so uh we 00:11:08.079 --> 00:11:10.110 align:start position:0% now it's just averaging out uh so uh we are<00:11:08.200> trying<00:11:08.480> to<00:11:08.880> take<00:11:09.160> the<00:11:09.320> probabilities<00:11:09.920> of 00:11:10.110 --> 00:11:10.120 align:start position:0% are trying to take the probabilities of 00:11:10.120 --> 00:11:13.670 align:start position:0% are trying to take the probabilities of each<00:11:10.279> and<00:11:10.519> every<00:11:11.480> uh<00:11:11.880> CH<00:11:12.160> children<00:11:12.480> ndes<00:11:13.480> uh 00:11:13.670 --> 00:11:13.680 align:start position:0% each and every uh CH children ndes uh 00:11:13.680 --> 00:11:18.069 align:start position:0% each and every uh CH children ndes uh and<00:11:13.920> then<00:11:14.800> uh<00:11:15.240> finding<00:11:16.240> uh<00:11:16.440> the<00:11:16.760> score<00:11:17.160> for<00:11:17.440> the 00:11:18.069 --> 00:11:18.079 align:start position:0% and then uh finding uh the score for the 00:11:18.079 --> 00:11:21.629 align:start position:0% and then uh finding uh the score for the parent<00:11:18.560> based<00:11:18.880> on<00:11:19.079> those<00:11:19.880> uh<00:11:20.360> scoring<00:11:21.360> uh<00:11:21.440> for 00:11:21.629 --> 00:11:21.639 align:start position:0% parent based on those uh scoring uh for 00:11:21.639 --> 00:11:23.870 align:start position:0% parent based on those uh scoring uh for the<00:11:21.760> children<00:11:22.120> so<00:11:22.519> yeah<00:11:22.720> that's 00:11:23.870 --> 00:11:23.880 align:start position:0% the children so yeah that's 00:11:23.880 --> 00:11:27.190 align:start position:0% the children so yeah that's the<00:11:24.880> uh<00:11:25.880> scoring<00:11:26.320> mechanism<00:11:26.760> we<00:11:26.880> are<00:11:26.959> trying 00:11:27.190 --> 00:11:27.200 align:start position:0% the uh scoring mechanism we are trying 00:11:27.200 --> 00:11:29.829 align:start position:0% the uh scoring mechanism we are trying to<00:11:27.360> work<00:11:27.560> on<00:11:28.440> and<00:11:28.920> uh 00:11:29.829 --> 00:11:29.839 align:start position:0% to work on and uh 00:11:29.839 --> 00:11:34.389 align:start position:0% to work on and uh I<00:11:29.920> mean<00:11:30.720> perfect<00:11:31.120> the<00:11:31.440> graph<00:11:32.440> oh<00:11:32.639> yeah<00:11:33.360> awesome 00:11:34.389 --> 00:11:34.399 align:start position:0% I mean perfect the graph oh yeah awesome 00:11:34.399 --> 00:11:37.509 align:start position:0% I mean perfect the graph oh yeah awesome uh<00:11:35.399> so<00:11:35.639> yeah<00:11:35.880> this<00:11:36.040> is<00:11:36.680> uh<00:11:36.800> what<00:11:36.920> I<00:11:37.040> was<00:11:37.279> talking 00:11:37.509 --> 00:11:37.519 align:start position:0% uh so yeah this is uh what I was talking 00:11:37.519 --> 00:11:39.629 align:start position:0% uh so yeah this is uh what I was talking about<00:11:38.000> where<00:11:38.160> we<00:11:38.279> were<00:11:38.519> just<00:11:38.720> randomly 00:11:39.629 --> 00:11:39.639 align:start position:0% about where we were just randomly 00:11:39.639 --> 00:11:42.110 align:start position:0% about where we were just randomly poisoning<00:11:40.160> the<00:11:40.279> nodes<00:11:40.959> uh<00:11:41.079> in<00:11:41.200> the<00:11:41.320> network 00:11:42.110 --> 00:11:42.120 align:start position:0% poisoning the nodes uh in the network 00:11:42.120 --> 00:11:44.389 align:start position:0% poisoning the nodes uh in the network you<00:11:42.360> don't<00:11:42.639> really<00:11:42.880> see<00:11:43.240> a<00:11:43.959> massive 00:11:44.389 --> 00:11:44.399 align:start position:0% you don't really see a massive 00:11:44.399 --> 00:11:47.590 align:start position:0% you don't really see a massive difference<00:11:44.800> between<00:11:45.279> rated<00:11:45.639> list<00:11:46.040> on<00:11:46.240> and<00:11:46.600> off 00:11:47.590 --> 00:11:47.600 align:start position:0% difference between rated list on and off 00:11:47.600 --> 00:11:51.069 align:start position:0% difference between rated list on and off uh<00:11:48.320> so<00:11:49.320> as<00:11:49.440> you<00:11:49.560> can<00:11:49.720> see<00:11:50.040> like<00:11:50.519> there's<00:11:50.760> not<00:11:50.920> a 00:11:51.069 --> 00:11:51.079 align:start position:0% uh so as you can see like there's not a 00:11:51.079 --> 00:11:52.670 align:start position:0% uh so as you can see like there's not a lot<00:11:51.240> of<00:11:51.399> difference<00:11:51.720> but<00:11:51.880> there's<00:11:52.040> a<00:11:52.320> small 00:11:52.670 --> 00:11:52.680 align:start position:0% lot of difference but there's a small 00:11:52.680 --> 00:11:57.069 align:start position:0% lot of difference but there's a small difference<00:11:53.160> but<00:11:53.760> uh 00:11:57.069 --> 00:11:57.079 align:start position:0% 00:11:57.079 --> 00:12:00.069 align:start position:0% still<00:11:58.079> yeah<00:11:58.560> this<00:11:58.680> is<00:11:58.880> where<00:11:59.040> we 00:12:00.069 --> 00:12:00.079 align:start position:0% still yeah this is where we 00:12:00.079 --> 00:12:04.310 align:start position:0% still yeah this is where we uh<00:12:00.680> shine<00:12:01.680> is<00:12:02.240> uh<00:12:02.440> the<00:12:02.839> rated<00:12:03.279> list<00:12:03.639> performs 00:12:04.310 --> 00:12:04.320 align:start position:0% uh shine is uh the rated list performs 00:12:04.320 --> 00:12:07.110 align:start position:0% uh shine is uh the rated list performs much<00:12:04.600> better<00:12:04.920> than<00:12:05.240> just<00:12:05.480> randomly<00:12:06.120> sampling 00:12:07.110 --> 00:12:07.120 align:start position:0% much better than just randomly sampling 00:12:07.120 --> 00:12:12.550 align:start position:0% much better than just randomly sampling uh<00:12:07.480> from<00:12:07.680> the<00:12:07.839> network<00:12:08.680> in<00:12:08.959> case<00:12:09.160> of<00:12:09.560> 90%<00:12:10.760> um<00:12:11.760> uh 00:12:12.550 --> 00:12:12.560 align:start position:0% uh from the network in case of 90% um uh 00:12:12.560 --> 00:12:15.470 align:start position:0% uh from the network in case of 90% um uh poisoning<00:12:13.120> so<00:12:13.399> this<00:12:13.519> is<00:12:13.760> where<00:12:14.320> there<00:12:14.480> was<00:12:14.639> a 00:12:15.470 --> 00:12:15.480 align:start position:0% poisoning so this is where there was a 00:12:15.480 --> 00:12:20.949 align:start position:0% poisoning so this is where there was a uh<00:12:15.959> entire<00:12:16.399> subtree<00:12:17.440> marked<00:12:18.560> uh<00:12:19.560> um<00:12:19.760> marked<00:12:20.360> as 00:12:20.949 --> 00:12:20.959 align:start position:0% uh entire subtree marked uh um marked as 00:12:20.959 --> 00:12:24.870 align:start position:0% uh entire subtree marked uh um marked as malicious<00:12:21.959> and<00:12:22.480> I<00:12:22.680> think<00:12:23.079> I<00:12:23.399> have<00:12:24.000> a 00:12:24.870 --> 00:12:24.880 align:start position:0% malicious and I think I have a 00:12:24.880 --> 00:12:28.670 align:start position:0% malicious and I think I have a small<00:12:25.880> yeah<00:12:26.199> so<00:12:27.000> the<00:12:27.279> those<00:12:27.399> are<00:12:27.639> the<00:12:28.320> sort<00:12:28.560> of 00:12:28.670 --> 00:12:28.680 align:start position:0% small yeah so the those are the sort of 00:12:28.680 --> 00:12:31.110 align:start position:0% small yeah so the those are the sort of future<00:12:29.199> what<00:12:29.320> we<00:12:29.399> are<00:12:29.480> trying<00:12:29.680> to<00:12:29.800> do<00:12:30.079> is<00:12:30.360> that 00:12:31.110 --> 00:12:31.120 align:start position:0% future what we are trying to do is that 00:12:31.120 --> 00:12:33.710 align:start position:0% future what we are trying to do is that uh<00:12:31.240> since<00:12:31.560> our<00:12:31.920> implementation<00:12:32.760> is<00:12:33.040> still<00:12:33.440> in 00:12:33.710 --> 00:12:33.720 align:start position:0% uh since our implementation is still in 00:12:33.720 --> 00:12:37.150 align:start position:0% uh since our implementation is still in like<00:12:33.920> super<00:12:34.680> native<00:12:35.680> uh<00:12:35.959> uh<00:12:36.399> like<00:12:36.600> primitive 00:12:37.150 --> 00:12:37.160 align:start position:0% like super native uh uh like primitive 00:12:37.160 --> 00:12:40.030 align:start position:0% like super native uh uh like primitive stage<00:12:37.480> right<00:12:37.639> now<00:12:37.920> so<00:12:38.519> uh<00:12:38.680> we<00:12:38.839> would<00:12:39.079> like<00:12:39.320> to 00:12:40.030 --> 00:12:40.040 align:start position:0% stage right now so uh we would like to 00:12:40.040 --> 00:12:43.910 align:start position:0% stage right now so uh we would like to know<00:12:40.320> Implement<00:12:41.240> uh<00:12:41.800> a<00:12:42.120> better<00:12:43.120> uh 00:12:43.910 --> 00:12:43.920 align:start position:0% know Implement uh a better uh 00:12:43.920 --> 00:12:46.389 align:start position:0% know Implement uh a better uh simulator<00:12:44.920> and 00:12:46.389 --> 00:12:46.399 align:start position:0% simulator and 00:12:46.399 --> 00:12:50.069 align:start position:0% simulator and uh<00:12:47.399> yeah<00:12:47.839> I<00:12:47.959> think<00:12:48.399> uh<00:12:48.600> this<00:12:49.160> uh<00:12:49.320> probably<00:12:49.680> sums 00:12:50.069 --> 00:12:50.079 align:start position:0% uh yeah I think uh this uh probably sums 00:12:50.079 --> 00:12:53.509 align:start position:0% uh yeah I think uh this uh probably sums up<00:12:51.040> our 00:12:53.509 --> 00:12:53.519 align:start position:0% up our 00:12:53.519 --> 00:12:55.069 align:start position:0% up our uh 00:12:55.069 --> 00:12:55.079 align:start position:0% uh 00:12:55.079 --> 00:13:01.470 align:start position:0% uh that's<00:12:56.160> yeah<00:12:57.160> yeah 00:13:01.470 --> 00:13:01.480 align:start position:0% 00:13:01.480 --> 00:13:03.430 align:start position:0% uh<00:13:01.880> so 00:13:03.430 --> 00:13:03.440 align:start position:0% uh so 00:13:03.440 --> 00:13:05.310 align:start position:0% uh so um<00:13:04.440> did 00:13:05.310 --> 00:13:05.320 align:start position:0% um did 00:13:05.320 --> 00:13:08.949 align:start position:0% um did we<00:13:06.320> can<00:13:06.399> you<00:13:06.560> go<00:13:06.680> back<00:13:06.800> to<00:13:06.880> the<00:13:07.519> graph<00:13:08.519> yeah<00:13:08.680> so 00:13:08.949 --> 00:13:08.959 align:start position:0% we can you go back to the graph yeah so 00:13:08.959 --> 00:13:10.350 align:start position:0% we can you go back to the graph yeah so just<00:13:09.079> want<00:13:09.240> to<00:13:09.440> point<00:13:09.639> out<00:13:09.839> that<00:13:10.000> all<00:13:10.160> these 00:13:10.350 --> 00:13:10.360 align:start position:0% just want to point out that all these 00:13:10.360 --> 00:13:13.189 align:start position:0% just want to point out that all these graphs<00:13:10.720> were<00:13:10.880> done<00:13:11.120> over<00:13:11.440> like<00:13:11.839> uh<00:13:12.079> 100<00:13:12.399> runs 00:13:13.189 --> 00:13:13.199 align:start position:0% graphs were done over like uh 100 runs 00:13:13.199 --> 00:13:17.470 align:start position:0% graphs were done over like uh 100 runs at<00:13:13.440> different<00:13:13.760> thresholds<00:13:15.000> sorry<00:13:16.000> um<00:13:16.839> so<00:13:17.160> all 00:13:17.470 --> 00:13:17.480 align:start position:0% at different thresholds sorry um so all 00:13:17.480 --> 00:13:19.870 align:start position:0% at different thresholds sorry um so all all<00:13:17.680> these<00:13:18.279> um<00:13:18.800> uh<00:13:18.920> metrics<00:13:19.320> were<00:13:19.519> plotted 00:13:19.870 --> 00:13:19.880 align:start position:0% all these um uh metrics were plotted 00:13:19.880 --> 00:13:22.230 align:start position:0% all these um uh metrics were plotted over<00:13:20.440> 100<00:13:20.720> runs<00:13:21.600> uh<00:13:21.800> with<00:13:21.959> different 00:13:22.230 --> 00:13:22.240 align:start position:0% over 100 runs uh with different 00:13:22.240 --> 00:13:24.030 align:start position:0% over 100 runs uh with different thresholds<00:13:22.800> and<00:13:22.959> different<00:13:23.240> strategies<00:13:23.720> TR 00:13:24.030 --> 00:13:24.040 align:start position:0% thresholds and different strategies TR 00:13:24.040 --> 00:13:26.990 align:start position:0% thresholds and different strategies TR out<00:13:24.720> and<00:13:24.920> for<00:13:25.120> the<00:13:25.199> rated<00:13:25.560> list<00:13:25.800> we<00:13:26.000> picked<00:13:26.760> the 00:13:26.990 --> 00:13:27.000 align:start position:0% out and for the rated list we picked the 00:13:27.000 --> 00:13:31.350 align:start position:0% out and for the rated list we picked the best<00:13:27.360> score<00:13:28.320> with<00:13:28.880> uh<00:13:29.199> in<00:13:29.480> all<00:13:30.480> among<00:13:30.839> all 00:13:31.350 --> 00:13:31.360 align:start position:0% best score with uh in all among all 00:13:31.360 --> 00:13:33.790 align:start position:0% best score with uh in all among all different<00:13:32.000> strategies<00:13:32.800> right<00:13:33.160> so<00:13:33.360> best<00:13:33.560> score 00:13:33.790 --> 00:13:33.800 align:start position:0% different strategies right so best score 00:13:33.800 --> 00:13:36.150 align:start position:0% different strategies right so best score first<00:13:34.120> average<00:13:34.399> score<00:13:34.639> first<00:13:34.959> and<00:13:35.160> everything 00:13:36.150 --> 00:13:36.160 align:start position:0% first average score first and everything 00:13:36.160 --> 00:13:38.750 align:start position:0% first average score first and everything uh<00:13:36.320> where<00:13:36.519> the<00:13:36.639> rated<00:13:36.959> list<00:13:37.240> off<00:13:37.560> is<00:13:37.800> just 00:13:38.750 --> 00:13:38.760 align:start position:0% uh where the rated list off is just 00:13:38.760 --> 00:13:41.389 align:start position:0% uh where the rated list off is just plain<00:13:39.079> old<00:13:39.399> naive<00:13:39.959> randomly<00:13:40.440> sampling<00:13:41.240> from 00:13:41.389 --> 00:13:41.399 align:start position:0% plain old naive randomly sampling from 00:13:41.399 --> 00:13:44.350 align:start position:0% plain old naive randomly sampling from the 00:13:44.350 --> 00:13:44.360 align:start position:0% 00:13:44.360 --> 00:13:48.389 align:start position:0% network<00:13:45.360> I<00:13:46.079> just<00:13:46.360> for<00:13:47.360> probably<00:13:47.800> just<00:13:47.959> for 00:13:48.389 --> 00:13:48.399 align:start position:0% network I just for probably just for 00:13:48.399 --> 00:13:49.509 align:start position:0% network I just for probably just for visual 00:13:49.509 --> 00:13:49.519 align:start position:0% visual 00:13:49.519 --> 00:13:52.829 align:start position:0% visual visualization<00:13:50.519> uh<00:13:51.000> this<00:13:51.199> is<00:13:51.600> the<00:13:52.360> sort<00:13:52.600> of 00:13:52.829 --> 00:13:52.839 align:start position:0% visualization uh this is the sort of 00:13:52.839 --> 00:13:57.430 align:start position:0% visualization uh this is the sort of graph<00:13:53.839> that<00:13:54.639> the<00:13:55.639> uh<00:13:55.839> big<00:13:56.360> blue<00:13:56.720> ball<00:13:57.160> is 00:13:57.430 --> 00:13:57.440 align:start position:0% graph that the uh big blue ball is 00:13:57.440 --> 00:13:59.670 align:start position:0% graph that the uh big blue ball is basically<00:13:57.839> a<00:13:58.000> network<00:13:58.480> but<00:13:58.680> then<00:13:59.279> it's<00:13:59.440> all 00:13:59.670 --> 00:13:59.680 align:start position:0% basically a network but then it's all 00:13:59.680 --> 00:14:01.509 align:start position:0% basically a network but then it's all pulled<00:14:00.000> out<00:14:00.320> and<00:14:00.639> those<00:14:00.800> are<00:14:00.959> the<00:14:01.079> malicious 00:14:01.509 --> 00:14:01.519 align:start position:0% pulled out and those are the malicious 00:14:01.519 --> 00:14:06.550 align:start position:0% pulled out and those are the malicious notes<00:14:01.920> yeah<00:14:02.759> so<00:14:03.199> yeah<00:14:03.399> thanks 00:14:06.550 --> 00:14:06.560 align:start position:0% 00:14:06.560 --> 00:14:09.030 align:start position:0% guys<00:14:07.560> all<00:14:07.720> right<00:14:07.959> any<00:14:08.199> questions<00:14:08.600> for<00:14:08.839> these 00:14:09.030 --> 00:14:09.040 align:start position:0% guys all right any questions for these 00:14:09.040 --> 00:14:18.350 align:start position:0% guys all right any questions for these two<00:14:09.360> on<00:14:09.560> the<00:14:09.680> rated<00:14:10.079> list 00:14:18.350 --> 00:14:18.360 align:start position:0% 00:14:18.360 --> 00:14:20.629 align:start position:0% project<00:14:19.360> uh<00:14:19.560> just<00:14:19.639> trying<00:14:19.839> to<00:14:20.199> understand<00:14:20.440> the 00:14:20.629 --> 00:14:20.639 align:start position:0% project uh just trying to understand the 00:14:20.639 --> 00:14:23.870 align:start position:0% project uh just trying to understand the Civil<00:14:21.399> graph<00:14:21.680> that<00:14:21.759> you<00:14:21.959> had<00:14:22.800> you<00:14:23.240> you 00:14:23.870 --> 00:14:23.880 align:start position:0% Civil graph that you had you you 00:14:23.880 --> 00:14:26.629 align:start position:0% Civil graph that you had you you injected<00:14:24.360> lots<00:14:24.560> of<00:14:24.800> Cil<00:14:25.199> nodes<00:14:25.600> into<00:14:25.880> the<00:14:26.160> tree 00:14:26.629 --> 00:14:26.639 align:start position:0% injected lots of Cil nodes into the tree 00:14:26.639 --> 00:14:28.189 align:start position:0% injected lots of Cil nodes into the tree and<00:14:27.000> how<00:14:27.079> do<00:14:27.199> you<00:14:27.320> determine<00:14:27.720> which<00:14:27.839> ones<00:14:28.000> were 00:14:28.189 --> 00:14:28.199 align:start position:0% and how do you determine which ones were 00:14:28.199 --> 00:14:29.749 align:start position:0% and how do you determine which ones were CBL<00:14:28.560> and<00:14:28.639> which<00:14:28.720> ones<00:14:28.880> wer 00:14:29.749 --> 00:14:29.759 align:start position:0% CBL and which ones wer 00:14:29.759 --> 00:14:32.350 align:start position:0% CBL and which ones wer or<00:14:29.880> have<00:14:30.000> I<00:14:30.560> misunderstood<00:14:31.560> um<00:14:32.000> are<00:14:32.160> you 00:14:32.350 --> 00:14:32.360 align:start position:0% or have I misunderstood um are you 00:14:32.360 --> 00:14:34.150 align:start position:0% or have I misunderstood um are you asking<00:14:32.639> about<00:14:32.839> the<00:14:32.959> implementation<00:14:33.680> detail 00:14:34.150 --> 00:14:34.160 align:start position:0% asking about the implementation detail 00:14:34.160 --> 00:14:37.870 align:start position:0% asking about the implementation detail or<00:14:34.480> like<00:14:35.000> uh<00:14:35.759> no<00:14:36.279> kind<00:14:36.399> of<00:14:36.519> in<00:14:36.639> general<00:14:36.959> so<00:14:37.440> like 00:14:37.870 --> 00:14:37.880 align:start position:0% or like uh no kind of in general so like 00:14:37.880 --> 00:14:42.509 align:start position:0% or like uh no kind of in general so like okay<00:14:38.360> so<00:14:39.120> um<00:14:39.440> a<00:14:39.560> civil<00:14:39.920> node<00:14:40.839> uh<00:14:41.839> would 00:14:42.509 --> 00:14:42.519 align:start position:0% okay so um a civil node uh would 00:14:42.519 --> 00:14:45.069 align:start position:0% okay so um a civil node uh would basically<00:14:43.000> not<00:14:43.279> reply<00:14:43.880> for<00:14:44.120> any<00:14:44.360> P<00:14:44.639> sampling 00:14:45.069 --> 00:14:45.079 align:start position:0% basically not reply for any P sampling 00:14:45.079 --> 00:14:48.509 align:start position:0% basically not reply for any P sampling requests<00:14:46.000> coming<00:14:46.279> to<00:14:46.480> it<00:14:46.959> right<00:14:47.880> uh<00:14:48.120> we<00:14:48.240> can 00:14:48.509 --> 00:14:48.519 align:start position:0% requests coming to it right uh we can 00:14:48.519 --> 00:14:50.829 align:start position:0% requests coming to it right uh we can definitely<00:14:49.040> test<00:14:49.440> for<00:14:49.839> more<00:14:50.160> cases<00:14:50.480> where<00:14:50.680> the 00:14:50.829 --> 00:14:50.839 align:start position:0% definitely test for more cases where the 00:14:50.839 --> 00:14:53.670 align:start position:0% definitely test for more cases where the attacks<00:14:51.120> are<00:14:51.279> more<00:14:51.759> nuanced<00:14:52.759> uh<00:14:52.920> as<00:14:53.120> in<00:14:53.320> it<00:14:53.519> can 00:14:53.670 --> 00:14:53.680 align:start position:0% attacks are more nuanced uh as in it can 00:14:53.680 --> 00:14:55.790 align:start position:0% attacks are more nuanced uh as in it can the<00:14:53.839> adversary<00:14:54.360> can<00:14:54.480> be<00:14:54.680> adaptive<00:14:55.519> uh<00:14:55.600> where 00:14:55.790 --> 00:14:55.800 align:start position:0% the adversary can be adaptive uh where 00:14:55.800 --> 00:14:58.550 align:start position:0% the adversary can be adaptive uh where the<00:14:55.880> Cil<00:14:56.199> node<00:14:56.720> only<00:14:57.040> responds<00:14:57.600> to<00:14:58.120> a<00:14:58.279> certain 00:14:58.550 --> 00:14:58.560 align:start position:0% the Cil node only responds to a certain 00:14:58.560 --> 00:15:01.829 align:start position:0% the Cil node only responds to a certain number<00:14:58.800> of<00:14:59.079> nodes<00:14:59.839> but<00:15:00.040> not<00:15:00.639> uh<00:15:01.160> other<00:15:01.639> other 00:15:01.829 --> 00:15:01.839 align:start position:0% number of nodes but not uh other other 00:15:01.839 --> 00:15:03.430 align:start position:0% number of nodes but not uh other other nodes<00:15:02.360> and<00:15:02.480> if<00:15:02.600> you<00:15:02.680> want<00:15:02.839> to<00:15:03.079> actually 00:15:03.430 --> 00:15:03.440 align:start position:0% nodes and if you want to actually 00:15:03.440 --> 00:15:07.470 align:start position:0% nodes and if you want to actually Eclipse<00:15:04.199> a<00:15:04.399> particular<00:15:05.320> uh<00:15:06.199> node<00:15:06.800> you<00:15:07.000> would 00:15:07.470 --> 00:15:07.480 align:start position:0% Eclipse a particular uh node you would 00:15:07.480 --> 00:15:09.670 align:start position:0% Eclipse a particular uh node you would you<00:15:07.560> would<00:15:07.800> flood<00:15:08.120> its<00:15:08.320> local<00:15:08.639> space<00:15:08.959> and<00:15:09.440> not 00:15:09.670 --> 00:15:09.680 align:start position:0% you would flood its local space and not 00:15:09.680 --> 00:15:11.629 align:start position:0% you would flood its local space and not just<00:15:09.880> reply<00:15:10.199> to<00:15:10.440> that<00:15:10.720> particular<00:15:11.120> node<00:15:11.480> but 00:15:11.629 --> 00:15:11.639 align:start position:0% just reply to that particular node but 00:15:11.639 --> 00:15:13.910 align:start position:0% just reply to that particular node but reply<00:15:11.959> to<00:15:12.120> every<00:15:12.320> other<00:15:12.480> node<00:15:13.240> so<00:15:13.519> yeah<00:15:13.720> so<00:15:13.839> you 00:15:13.910 --> 00:15:13.920 align:start position:0% reply to every other node so yeah so you 00:15:13.920 --> 00:15:15.790 align:start position:0% reply to every other node so yeah so you were<00:15:14.120> detecting<00:15:14.920> the<00:15:15.040> sibl<00:15:15.399> you<00:15:15.480> didn't<00:15:15.639> just 00:15:15.790 --> 00:15:15.800 align:start position:0% were detecting the sibl you didn't just 00:15:15.800 --> 00:15:17.189 align:start position:0% were detecting the sibl you didn't just assume<00:15:16.079> that<00:15:16.199> they<00:15:16.279> were<00:15:16.440> sibl<00:15:16.920> right<00:15:17.079> you 00:15:17.189 --> 00:15:17.199 align:start position:0% assume that they were sibl right you 00:15:17.199 --> 00:15:18.710 align:start position:0% assume that they were sibl right you didn't<00:15:17.399> just<00:15:17.560> say<00:15:17.800> these<00:15:17.959> ones<00:15:18.120> are<00:15:18.279> malicious 00:15:18.710 --> 00:15:18.720 align:start position:0% didn't just say these ones are malicious 00:15:18.720 --> 00:15:20.870 align:start position:0% didn't just say these ones are malicious you<00:15:18.880> actually<00:15:19.639> based<00:15:19.880> on<00:15:20.040> their<00:15:20.240> responses 00:15:20.870 --> 00:15:20.880 align:start position:0% you actually based on their responses 00:15:20.880 --> 00:15:23.189 align:start position:0% you actually based on their responses you<00:15:21.040> determined<00:15:21.440> that<00:15:21.560> they<00:15:21.639> were<00:15:22.480> attacks 00:15:23.189 --> 00:15:23.199 align:start position:0% you determined that they were attacks 00:15:23.199 --> 00:15:25.310 align:start position:0% you determined that they were attacks yes 00:15:25.310 --> 00:15:25.320 align:start position:0% yes 00:15:25.320 --> 00:15:28.670 align:start position:0% yes yeah<00:15:26.320> yeah<00:15:26.560> the<00:15:27.399> parent<00:15:27.759> child<00:15:28.079> relationship 00:15:28.670 --> 00:15:28.680 align:start position:0% yeah yeah the parent child relationship 00:15:28.680 --> 00:15:30.550 align:start position:0% yeah yeah the parent child relationship that<00:15:28.800> you<00:15:29.040> you've<00:15:29.480> showed<00:15:29.800> in<00:15:29.920> the<00:15:30.040> graph<00:15:30.360> like 00:15:30.550 --> 00:15:30.560 align:start position:0% that you you've showed in the graph like 00:15:30.560 --> 00:15:33.470 align:start position:0% that you you've showed in the graph like how<00:15:30.720> do<00:15:30.880> you<00:15:31.240> come<00:15:31.440> up<00:15:31.639> with<00:15:31.800> the<00:15:32.279> peer<00:15:33.160> parent 00:15:33.470 --> 00:15:33.480 align:start position:0% how do you come up with the peer parent 00:15:33.480 --> 00:15:36.389 align:start position:0% how do you come up with the peer parent child<00:15:33.759> relationship<00:15:34.319> is<00:15:34.399> it<00:15:34.639> like<00:15:35.160> if<00:15:35.560> this 00:15:36.389 --> 00:15:36.399 align:start position:0% child relationship is it like if this 00:15:36.399 --> 00:15:39.350 align:start position:0% child relationship is it like if this peer<00:15:37.399> like<00:15:37.920> told<00:15:38.240> you<00:15:38.399> about<00:15:38.600> the<00:15:38.720> other<00:15:38.959> peer 00:15:39.350 --> 00:15:39.360 align:start position:0% peer like told you about the other peer 00:15:39.360 --> 00:15:42.870 align:start position:0% peer like told you about the other peer then<00:15:39.639> it's<00:15:40.079> its<00:15:40.440> child<00:15:41.440> yes<00:15:41.759> so<00:15:42.000> it's<00:15:42.199> a<00:15:42.399> it's 00:15:42.870 --> 00:15:42.880 align:start position:0% then it's its child yes so it's a it's 00:15:42.880 --> 00:15:45.030 align:start position:0% then it's its child yes so it's a it's much<00:15:43.040> more<00:15:43.199> of<00:15:43.319> a<00:15:43.480> local<00:15:43.800> view<00:15:44.560> rather<00:15:44.759> than<00:15:44.880> a 00:15:45.030 --> 00:15:45.040 align:start position:0% much more of a local view rather than a 00:15:45.040 --> 00:15:47.309 align:start position:0% much more of a local view rather than a global<00:15:45.319> view<00:15:45.639> where<00:15:46.240> we<00:15:46.440> it's<00:15:46.839> basically<00:15:47.199> just 00:15:47.309 --> 00:15:47.319 align:start position:0% global view where we it's basically just 00:15:47.319 --> 00:15:49.790 align:start position:0% global view where we it's basically just like<00:15:47.480> peer<00:15:47.720> exchange<00:15:48.079> where<00:15:48.240> we<00:15:48.440> just<00:15:48.720> ask<00:15:49.000> for 00:15:49.790 --> 00:15:49.800 align:start position:0% like peer exchange where we just ask for 00:15:49.800 --> 00:15:51.749 align:start position:0% like peer exchange where we just ask for uh<00:15:49.959> peers<00:15:50.360> from<00:15:50.639> particular<00:15:51.040> node<00:15:51.480> like<00:15:51.600> so 00:15:51.749 --> 00:15:51.759 align:start position:0% uh peers from particular node like so 00:15:51.759 --> 00:15:53.509 align:start position:0% uh peers from particular node like so first<00:15:51.959> I<00:15:52.120> asked<00:15:52.319> from<00:15:52.480> the<00:15:52.680> nodes<00:15:53.160> that<00:15:53.279> I<00:15:53.360> am 00:15:53.509 --> 00:15:53.519 align:start position:0% first I asked from the nodes that I am 00:15:53.519 --> 00:15:57.110 align:start position:0% first I asked from the nodes that I am connected<00:15:53.920> to<00:15:54.480> for<00:15:54.759> their<00:15:54.959> peers<00:15:55.639> and<00:15:55.839> then<00:15:56.600> I 00:15:57.110 --> 00:15:57.120 align:start position:0% connected to for their peers and then I 00:15:57.120 --> 00:15:59.550 align:start position:0% connected to for their peers and then I make<00:15:57.319> a<00:15:57.440> list<00:15:57.680> of<00:15:57.959> those<00:15:58.160> peers<00:15:58.519> and<00:15:58.680> ask<00:15:59.319> ask 00:15:59.550 --> 00:15:59.560 align:start position:0% make a list of those peers and ask ask 00:15:59.560 --> 00:16:02.550 align:start position:0% make a list of those peers and ask ask for<00:15:59.720> peers<00:16:00.079> from<00:16:00.399> them<00:16:00.920> as<00:16:01.120> well<00:16:01.680> I<00:16:01.800> can<00:16:02.160> cap 00:16:02.550 --> 00:16:02.560 align:start position:0% for peers from them as well I can cap 00:16:02.560 --> 00:16:05.350 align:start position:0% for peers from them as well I can cap this<00:16:03.319> uh<00:16:03.759> basically<00:16:04.240> we<00:16:04.480> in<00:16:04.600> the<00:16:04.759> rated<00:16:05.079> list 00:16:05.350 --> 00:16:05.360 align:start position:0% this uh basically we in the rated list 00:16:05.360 --> 00:16:08.710 align:start position:0% this uh basically we in the rated list we<00:16:05.519> have<00:16:05.720> it<00:16:05.959> capped<00:16:06.279> at<00:16:06.639> 100<00:16:07.639> uh<00:16:07.759> Max<00:16:08.120> children 00:16:08.710 --> 00:16:08.720 align:start position:0% we have it capped at 100 uh Max children 00:16:08.720 --> 00:16:12.150 align:start position:0% we have it capped at 100 uh Max children per<00:16:08.920> Noe<00:16:09.759> so<00:16:10.600> any<00:16:10.800> node<00:16:11.120> could<00:16:11.480> like<00:16:11.759> randomly 00:16:12.150 --> 00:16:12.160 align:start position:0% per Noe so any node could like randomly 00:16:12.160 --> 00:16:14.590 align:start position:0% per Noe so any node could like randomly sent<00:16:12.440> for<00:16:12.680> every<00:16:12.880> slot<00:16:13.720> randomly<00:16:14.160> or<00:16:14.319> maybe 00:16:14.590 --> 00:16:14.600 align:start position:0% sent for every slot randomly or maybe 00:16:14.600 --> 00:16:16.949 align:start position:0% sent for every slot randomly or maybe probabilistically<00:16:15.360> send<00:16:16.120> a<00:16:16.319> certain<00:16:16.600> set<00:16:16.800> of 00:16:16.949 --> 00:16:16.959 align:start position:0% probabilistically send a certain set of 00:16:16.959 --> 00:16:19.269 align:start position:0% probabilistically send a certain set of 100<00:16:17.199> nodes<00:16:17.920> to 00:16:19.269 --> 00:16:19.279 align:start position:0% 100 nodes to 00:16:19.279 --> 00:16:22.389 align:start position:0% 100 nodes to us<00:16:20.279> is<00:16:20.440> is<00:16:20.519> this<00:16:20.680> a<00:16:20.839> candidate<00:16:21.319> for<00:16:22.160> like 00:16:22.389 --> 00:16:22.399 align:start position:0% us is is this a candidate for like 00:16:22.399 --> 00:16:24.230 align:start position:0% us is is this a candidate for like replacing<00:16:23.040> CAD<00:16:23.240> based 00:16:24.230 --> 00:16:24.240 align:start position:0% replacing CAD based 00:16:24.240 --> 00:16:27.189 align:start position:0% replacing CAD based Discovery<00:16:25.240> it<00:16:25.399> wouldn't<00:16:25.839> replace<00:16:26.440> CAD<00:16:26.720> based 00:16:27.189 --> 00:16:27.199 align:start position:0% Discovery it wouldn't replace CAD based 00:16:27.199 --> 00:16:29.590 align:start position:0% Discovery it wouldn't replace CAD based Discovery<00:16:27.519> per<00:16:27.759> se<00:16:28.199> it<00:16:28.319> would<00:16:28.639> so<00:16:29.040> rated<00:16:29.319> list 00:16:29.590 --> 00:16:29.600 align:start position:0% Discovery per se it would so rated list 00:16:29.600 --> 00:16:31.949 align:start position:0% Discovery per se it would so rated list assumes<00:16:30.120> that<00:16:30.240> you<00:16:30.399> already<00:16:30.680> have<00:16:30.800> a<00:16:31.079> P2P 00:16:31.949 --> 00:16:31.959 align:start position:0% assumes that you already have a P2P 00:16:31.959 --> 00:16:35.629 align:start position:0% assumes that you already have a P2P network<00:16:32.959> uh<00:16:33.480> so<00:16:33.959> and<00:16:34.160> for<00:16:34.680> Discovery<00:16:35.120> it's<00:16:35.360> not 00:16:35.629 --> 00:16:35.639 align:start position:0% network uh so and for Discovery it's not 00:16:35.639 --> 00:16:37.269 align:start position:0% network uh so and for Discovery it's not really<00:16:36.240> like<00:16:36.360> you<00:16:36.480> would<00:16:36.600> need<00:16:36.880> bootstrap 00:16:37.269 --> 00:16:37.279 align:start position:0% really like you would need bootstrap 00:16:37.279 --> 00:16:38.910 align:start position:0% really like you would need bootstrap nodes<00:16:37.519> to<00:16:37.680> actually<00:16:37.880> form<00:16:38.199> a<00:16:38.440> this<00:16:38.600> thing<00:16:38.759> so 00:16:38.910 --> 00:16:38.920 align:start position:0% nodes to actually form a this thing so 00:16:38.920 --> 00:16:40.710 align:start position:0% nodes to actually form a this thing so it<00:16:39.040> becomes<00:16:39.319> a<00:16:39.440> little<00:16:39.920> more<00:16:40.199> complicated 00:16:40.710 --> 00:16:40.720 align:start position:0% it becomes a little more complicated 00:16:40.720 --> 00:16:43.949 align:start position:0% it becomes a little more complicated over<00:16:40.959> there<00:16:41.880> uh<00:16:42.160> but<00:16:42.399> yeah<00:16:43.279> but<00:16:43.519> what<00:16:43.680> we<00:16:43.800> are 00:16:43.949 --> 00:16:43.959 align:start position:0% over there uh but yeah but what we are 00:16:43.959 --> 00:16:46.389 align:start position:0% over there uh but yeah but what we are definitely<00:16:44.399> planning<00:16:44.720> on<00:16:45.040> is<00:16:46.000> like<00:16:46.160> if<00:16:46.240> you 00:16:46.389 --> 00:16:46.399 align:start position:0% definitely planning on is like if you 00:16:46.399 --> 00:16:50.230 align:start position:0% definitely planning on is like if you think<00:16:46.600> about<00:16:46.800> it<00:16:47.600> uh<00:16:47.800> per<00:16:48.120> slot<00:16:48.680> scores<00:16:49.680> uh 00:16:50.230 --> 00:16:50.240 align:start position:0% think about it uh per slot scores uh 00:16:50.240 --> 00:16:53.910 align:start position:0% think about it uh per slot scores uh they<00:16:50.440> they<00:16:50.560> make<00:16:50.839> sense<00:16:51.279> if<00:16:52.120> uh<00:16:52.480> you<00:16:52.920> assume 00:16:53.910 --> 00:16:53.920 align:start position:0% they they make sense if uh you assume 00:16:53.920 --> 00:16:56.150 align:start position:0% they they make sense if uh you assume that<00:16:54.519> uh<00:16:54.959> malicious<00:16:55.399> nodes<00:16:55.639> are<00:16:55.759> honest 00:16:56.150 --> 00:16:56.160 align:start position:0% that uh malicious nodes are honest 00:16:56.160 --> 00:16:58.590 align:start position:0% that uh malicious nodes are honest acting<00:16:56.600> until<00:16:56.920> that<00:16:57.120> slot<00:16:57.920> but<00:16:58.120> then<00:16:58.319> you<00:16:58.440> can 00:16:58.590 --> 00:16:58.600 align:start position:0% acting until that slot but then you can 00:16:58.600 --> 00:17:00.870 align:start position:0% acting until that slot but then you can also<00:16:59.040> have<00:16:59.199> malicious<00:16:59.680> nodes<00:17:00.360> that<00:17:00.560> start 00:17:00.870 --> 00:17:00.880 align:start position:0% also have malicious nodes that start 00:17:00.880 --> 00:17:03.269 align:start position:0% also have malicious nodes that start acting<00:17:01.160> honest<00:17:01.639> after<00:17:01.959> that<00:17:02.120> slot<00:17:02.800> and<00:17:02.920> a<00:17:03.040> per 00:17:03.269 --> 00:17:03.279 align:start position:0% acting honest after that slot and a per 00:17:03.279 --> 00:17:04.949 align:start position:0% acting honest after that slot and a per slot<00:17:03.519> score<00:17:03.759> wouldn't<00:17:04.079> really<00:17:04.360> help<00:17:04.640> at<00:17:04.799> that 00:17:04.949 --> 00:17:04.959 align:start position:0% slot score wouldn't really help at that 00:17:04.959 --> 00:17:08.150 align:start position:0% slot score wouldn't really help at that point<00:17:05.640> so<00:17:06.280> uh<00:17:06.480> but<00:17:07.079> the<00:17:07.360> the<00:17:07.520> good<00:17:07.720> thing<00:17:07.880> is 00:17:08.150 --> 00:17:08.160 align:start position:0% point so uh but the the good thing is 00:17:08.160 --> 00:17:11.189 align:start position:0% point so uh but the the good thing is the<00:17:08.760> uh<00:17:09.000> version<00:17:09.319> 1.2<00:17:10.039> gossip<00:17:10.400> scoring<00:17:11.079> uh 00:17:11.189 --> 00:17:11.199 align:start position:0% the uh version 1.2 gossip scoring uh 00:17:11.199 --> 00:17:12.750 align:start position:0% the uh version 1.2 gossip scoring uh provides<00:17:11.600> like<00:17:11.760> these<00:17:11.959> application<00:17:12.400> scores 00:17:12.750 --> 00:17:12.760 align:start position:0% provides like these application scores 00:17:12.760 --> 00:17:14.270 align:start position:0% provides like these application scores that<00:17:12.880> you<00:17:12.959> can<00:17:13.120> inject<00:17:13.439> into<00:17:13.679> them<00:17:14.039> and<00:17:14.160> they 00:17:14.270 --> 00:17:14.280 align:start position:0% that you can inject into them and they 00:17:14.280 --> 00:17:16.309 align:start position:0% that you can inject into them and they are<00:17:14.439> persisted<00:17:15.120> with<00:17:15.280> DK<00:17:15.679> parameters<00:17:16.120> and 00:17:16.309 --> 00:17:16.319 align:start position:0% are persisted with DK parameters and 00:17:16.319 --> 00:17:18.590 align:start position:0% are persisted with DK parameters and everything<00:17:16.640> so<00:17:16.799> we<00:17:17.319> plan<00:17:17.600> on<00:17:17.799> just<00:17:18.079> injecting 00:17:18.590 --> 00:17:18.600 align:start position:0% everything so we plan on just injecting 00:17:18.600 --> 00:17:21.429 align:start position:0% everything so we plan on just injecting the<00:17:18.720> rated<00:17:19.039> list<00:17:19.280> score<00:17:19.799> into<00:17:20.160> the<00:17:20.799> uh<00:17:21.039> Gossip 00:17:21.429 --> 00:17:21.439 align:start position:0% the rated list score into the uh Gossip 00:17:21.439 --> 00:17:26.710 align:start position:0% the rated list score into the uh Gossip V2<00:17:22.240> v1.2<00:17:23.240> scoring<00:17:24.679> yeah<00:17:25.679> all<00:17:25.839> right<00:17:26.120> thank<00:17:26.319> you 00:17:26.710 --> 00:17:26.720 align:start position:0% V2 v1.2 scoring yeah all right thank you 00:17:26.720 --> 00:17:28.350 align:start position:0% V2 v1.2 scoring yeah all right thank you let's<00:17:26.919> give<00:17:27.079> it<00:17:27.199> up<00:17:27.319> for<00:17:27.480> these<00:17:27.720> guys<00:17:27.959> one<00:17:28.079> more 00:17:28.350 --> 00:17:28.360 align:start position:0% let's give it up for these guys one more 00:17:28.360 --> 00:17:30.840 align:start position:0% let's give it up for these guys one more time