Analyzing network lifetime of wireless sensor networks with mathematical programming

According to the needs of typical applications, wireless sensor nodes are designed to be low-cost, small-sized, and energy-efficient devices. In order to satisfy these production requirements, they generally have scarce resources like energy, bandwidth, communication range, processing power, and memory. After the limitations related to cost and hardware are met by the progress on semiconductor, network, and materials technologies; energy becomes the most critical resource for Wireless Sensor Networks (WSNs). So, recent studies about WSNs concentrated on developing energy-efficient solutions to optimize network lifetime. However, most of these studies ignored the effect of finite bandwidth and discretization of transmission power on WSNs. On the other hand, there are different types of transmission power assignment strategies studied so far, and while it is evident that more fine-tuned power assignment improves network lifetime, the net impact of these strategies remains unclear. In this study, we develop novel mathematical programming frameworks which enable us not only to examine the effects of limited bandwidth and discrete transmission power control, but also to quantify the impact and make a systematic comparison of various power assignment strategies. We analyze the network bandwidth from several aspects with various system parameters. In order to obtain optimal network lifetime with specific parameters, we expose the methodology to determine the minimum amount of required bandwidth. We also investigate the effects of the granularity of power levels on energy dissipation characteristics. Different types of power assignment strategies are analyzed by using two sets of experimental data to compare the performance of these strategies in terms of network lifetime and link bandwidth. In order to see the effects of probabilistic radio propagation, widely used log-normal shadowing path loss model is adopted into existing models. Our results show that, link bandwidth affects network lifetime within a specific range. As interference rises, the amount of required bandwidth to obtain optimum lifetime increases as well. The granularity of discrete energy consumption has a profound impact on WSN lifetime and when discretization level ascends, network lifetime generally lessens. Results expose that while more fine-grained control of transmission power improves network lifetime, it also requires much more bandwidth.