4.2.4. Experiment Four

Aim: The next set of experiments was conducted to measure the efficiency of the proposed model compared to traditional recovery methods concerning the degree of complexity of the chosen strategies. In this case, the degree of complexity of the strategies was divided into three levels: low, medium, and high. The difficulty here is measured through utility functions. In the case of the log file size parameter, the range of size changes from small to medium to large is a measure of the complexity of the strategy. The same is the case for the other parameters such as handoff rate and mobility rate. Herein, the evaluation is based on the total payoff value that is calculated as the sum of the three utility functions' outputs.

Main Results: The results in Figure 7 confirm the superiority of the suggested model. As previously stated, since the proposed model is based on a knowledge base that was created after the pre-implementation of each chosen recovery protocol in various simulated environments, it automatically picks the most appropriate recovery procedure for the present situation.

Discussion: As expected, the log managemen<sup>t</sup> algorithm is preferred in the small area. However, this decision turns out to be unfavorable with a large log size, especially in distant regions, because the cost of transfer the log file becomes high, and thus the cost of recovery becomes more problematic. On the other hand, the hybrid method gave a good payoff compared to other algorithms whenever there was a multiplicity in the regions because it had taken the recovery point once. The same is true for agent-based recovery, as it is easy to find the recovery location by tracing the MH ids.

**Figure 7.** Comparison results in terms of total payoff.
