Performance Analysis Of Large Scale Wireless Sensor Networks

With the rapid development and increasing need of wireless sensor networks, theasymptotic analysis of large scale wireless sensor networks are getting more and moreattention from researchers and the networking community. Wireless sensor networkshavesomeuniquefeaturesandrequirements, suchasdynamicnetworkcondition, com-plex environment, connectivity and data integrity, which make the design and analysisof the network quite difcult and complex. In order to improve the network perfor-mance, many techniques or schemes, such as mobility, multicast, helping nodes andadding infrastructure are introduced and explored in literatures. Although challenging,the performance of the network shows some interesting characteristics in such scenar-ios. In this paper, we investigate the topology, capacity and delay performance of theheterogeneous or hybrid and wireless sensor networks. We study the network perfor-mance in diferent transmission mode and scenarios. We also explore methods anddesigns to improve the performance of wireless sensor networks while considering itsspecial characteristics and requirements.In this paper, we frst consider the heterogeneous wireless sensor networks andinvestigate how network topology infuences the network performance. For networkswithoutobstacles,weexplorethethroughputcapacityofnetworksdiferentnodedistri-butions. Ourresultsleadtosomeusefulguidelinesonoptimizingthenetworktopologyforsuchnetworkenvironments. Wefndthatregularhelpingnodesarethroughputopti-mal for network with uniformly distributed sensor nodes. If sensors are not uniformlydistributed, regular helping nodes are no longer the best choice. However, it is stillorder optimal for the network throughput capacity in the sense of scaling law.In practice, wireless sensor networks are often deployed in areas with hostile con- ditions. Obstacles distributed in these complex environments may have a huge impacton the distribution of sensors and the transmission of electromagnetic waves. In orderto deal with such complex network environments, we have to take all the above men-tioned dimensions into account and have a joint analysis of the throughput capacity.For networks with obstacles, we propose an optimization algorithm to explore a bettersensor distribution and improve the network capacity for given obstacles distribution.The algorithm complexity and performance are also analyzed to illustrate its featuresand efects from both quantitative and qualitative views.For hybrid wireless sensor networks, we investigate the impact of nodes mobilityand conclude how mobility improves tradeofs. We frst investigate the2-hop relay al-gorithm without redundancy in i.i.d. mobile networks. Then we introduce redundancyto improve the performance of the algorithm. Besides the i.i.d. mobility model, wealso consider Markovian dynamics to model users’mobility, such as random walkmobility model, which makes the results more realistic. When the number of base s-tations or the scale of multicast is large enough, our results show that infrastructuremode is better than ad-hoc mode for both of the algorithms. Our results also indicatethat when using base station to improve the performance of the network, investmentsmust be large enough to ensure sufcient amount of base stations, or it can even harmthe network’s performance.