Joint Design
Having described individual problems of design in mesh networks in an earlier chapter, we now describe design problems that more recent literature has focused on, which address the problem of jointly designing more than one functional or operational aspec
- PDF / 624,819 Bytes
- 36 Pages / 439.37 x 666.142 pts Page_size
- 68 Downloads / 198 Views
Joint Design
5.1 Joint Design As we discussed earlier, it is readily apparent that various individual design problems are themselves highly interdependent; this is well understood among researchers. Over the course of several years of research, it has become obvious that dealing with these interdependent problems jointly is preferable (indeed, almost unavoidable) in optimizing performance. For example, even a highly effective link scheduling algorithm may not yield very good performance, because some links exist in the network that are heavily used and also centrally located so that they interfere with many other links. It is attractive to reach into the solution of the routing problem, and adjust routes so that such links are less utilized, and therefore need to be less frequently scheduled. This gives rise to an area of designing algorithms that jointly provide both link schedules and routes. Previous surveys of literature largely pre-date this recent body of literature. Some surveys such as those by Foukalas et al. (2008) and Shariat et al. (2009) cover crosslayer design proposals but they focus on single-hop infrastructure networks only. We have ourselves surveyed this area previously (Pathak and Dutta 2010). In this chapter, we discuss the research problems that have emerged in this joint design area, and have been addressed in literature, with a brief survey of the literature. For a fuller survey, we refer the reader to Pathak and Dutta (2010).
5.2 Power Control and Scheduling As we have seen before, scheduling algorithms must take into consideration the interference relationships between the links which in turn is decided by the power assignments at nodes. The nodes transmitting at high power level creates higher interference links which reduces the overall spatial reuse when scheduled. One of the first solutions to problem of jointly scheduling and assigning power control, with P. H. Pathak and R. Dutta, Designing for Network and Service Continuity in Wireless Mesh Networks, Signals and Communication Technology, DOI: 10.1007/978-1-4614-4627-9_5, © Springer Science+Business Media New York 2013
133
134
5 Joint Design
the objective of maximizing throughput and minimizing the power consumption, was provided by ElBatt and Ephremides (2004). They provided a two-phase algorithm that is centralized and needs to be executed before every slot. In the first phase, the algorithm determines the maximum set of nodes that can transmit in a given slot with the constraint that they should be spatially separated by at least some distance to avoid mutual interference. In the second phase, such feasible sets of transmitting nodes are assigned power levels to meet their SINR constraints. Similarly, Chen and Lee (2006) propose a two phase distributed algorithm for power control and link scheduling in wireless networks with the objective of throughput enhancement by lowering interference. In the first phase, all nodes having data to send first probe the channel with some initial predetermined power by sending probe packets
Data Loading...
