On Capacity Conservation in Wireless Networks

Abstract

Spectrum allocation and application requirements often result in severe bandwidth limitations on wireless channels. A fundamental challenge is using the limited channel capacity in the most efficient way. This dissertation presents several capacityconservation schemes for wireless networks. The proposed algorithms span across the entire protocol stack but share the same objective of eliminatingredundancies and ensuring the timeliness of transmissions.At the lowest level, capacity conservation is improved with a header compression algorithm that unlike previous protocols is capable of compressing MAC headers on multiple-access (shared) channels. Previous header compression schemes are unable to compress MAC headers as they rely on link layer addressing for the correct interpretation of compressed fields.An adaptive encapsulation scheme is presented to maximize the number IP packets sent during a single contention for the channel in an effort to amortizethe overhead over multiple IP packets. Efficiency is improved by deferring transmissions to accumulate several packets, all while ensuring transport-leveldelay requirements and accounting for the time-varying nature of the transmission overhead as network conditions change. The result is anoutstanding improvement in the effective throughput.Another header compression algorithm is presented which leverages on routing information to minimize the overhead of frequent context initializations during periods of high node mobility. A statelesscompression scheme for control messages is also presented and is used to reduce the overhead of frequent routing updates.Finally, by addressing the redundancy and timeliness issues of multi-hop wireless broadcasting, we improve the capacity conservation in one of the most important applications in multi-hop wirelessenvironments.