The Research On Optimization Of Energy-aware Based Cache Replacement Strategy

In recent years, because of self-organization, dynamic and reliable characteristics,wireless sensor networks(WSN) are more widely used in various fields of environmental monitoring, industrial and agricultural production. Endless optimizations of sensor nodes are undertaken in order to overcome the limit of energy,computing power and storage space, which makes wireless sensor networks become one of the research hotspots in the field of network communications. Moreover, as a key technology for information sharing and distribution, caching technology can reduce network traffic and access latency, save energy, improve the overall performance of the wireless sensor network system.Firstly, In order to obtain the energy consumption distributed of sensor nodes under different working conditions, this paper makes in-depth analysis on the energy dissipation capacity of wireless sensor networks and sensor nodes, as well as dynamic power management(DPM) strategies. Secondly, this paper gives the research in wireless sensor network under the maintenance cost restrictions of weak cache consistency TTL mechanism, using directional diffusion routing protocol to establish a data transmission path from the sink node to the data source node. Then introduces the basic principles and the main idea of caching on cache placement policies and cache replacement strategy, and clarifies demand caching placement strategy based on dichotomy ideological that employed in this paper, as well as the research thought of replacement cost used in cache replacement of wireless sensor networks. Thirdly, this paper improve the BESS cache replacement algorithm of pursuing energy saving andthe LIRS cache replacement algorithm of pursuing high cache hit, analyzing the role that each factor constituting weight function plays in the algorithms. Then refine the factors that can achieve the coexistence of multi-objective optimization of wireless sensor networks, put the cache replacement algorithm with the coexistence of low energy consumption and high cache hit.Lastly, the simulation results show that, compared to the BESS algorithm and the LIRS algorithm, new algorithm decreased 15.07% ~ 23.18% and 11.01% ~ 18.46% in average latency, 4.71% ~ 8.95% and 20.31% ~ 21.36% on average energy consumption, and 5.33% ~ 12.70% and 1.71%- 4.94% on cache hit ratio.