摘要:Any node can receive a lot of information about the network by placing its interface into promiscuous mode. The information the node can receive can be used to build trust levels for different modes.
at is averaged over all the packets properly delivered for each of the data streams and for all the mobile scenarios. Therefore, there is an averaged value for each combination of:
a. Data traffic rate
b. MPR Coverage parameter
c. TC Redundancy parameter
4.4 Experimental Results:
Simulation work was performed as described in the previous sections. In this section the corresponding results are shown.
4.4.1 Scenarios without Data Traffic
The initial simulation work which was performed over static scenarios wanted to achieve some basic understanding about protocol’s performance and to get some insights on the effects of each proposed
strategy to increase the topology knowledge.
Some example tables are graphs are given below which will allow us to have some insight into the network and performance of OLSR in different situations.
Table 4-3: Percentages of nodes selected as MPRs for different values of MPR and TC
Table 4-3 and its corresponding graph, Graph 4-1, show how the amount of nodes selected as MPRs increase with the MPR parameter. Also, it is possible to notice that the amount of chosen MPRs is not affected by the TC strategy.
Graph 4-1: Percentage of nodes selected as MPRs for different values of MPRs and TC
4.4.2 Static Scenarios with Data Traffic
In the previous section, no data traffic was sent and all the scenarios were static, therefore, it is possible to assume that at some point in time the network reaches an stability state where the topology does not change, the nodes that were chosen as MPRs do not change their status and, for the same reason, the topology knowledge does not change either. Therefore, if that is true, what has to be examined is what the impact of data traffic. With that aim one single scenario was chosen and all the different strategies and traffic rates were applied to it while keeping track second by second of the Topology Knowledge and the percentage of nodes chosen as MPRs.
Graph 4-2 shows how the topology knowledge dramatically decreases when data traffic is injected. The topology knowledge drop is at second 35 which means that the last set of broadcasted TC messages properly received was at second number 20, right before the data sources started sending traffic. The last because the protocol configuration says that TC message information has to be kept as valid for up to TOP_HOLD_TIME=15 seconds if no more information is received. Therefore, the lost of TC messages due to high traffic load is reflected with some delay as a decrease on the topology knowledge.
Once that the traffic load decreases the topology knowledge increases again. On the other hand, the traffic load also originates loses in terms of Hello messages, these loses are reflected as an increase on the number of MPRs (Graph 4-3).
Graph 4-2: Topology knowledge for MPR1 under high traffic
Graph 4-3: Percentage of MPRs for MPR=1 and TC=2 under high traffic
Finally, the last metric that tells about protocol performance is the data delivery rate and it is shown in Graph 4-4. In this table we can clearly observe that the data delivery rate decreases with the traffic load going from 98% to 25% approx. Also the largest difference between every strategy combination, under the same tr
本论文由英语论文网提供整理,提供论文代写,英语论文代写,代写论文,代写英语论文,代写留学生论文,代写英文论文,留学生论文代写相关核心关键词搜索。