Distributed power control in asymmetric interference-limited networks

Power control in wireless communication networks reduces interference by regulating access to the shared wireless medium and enhances the overall network capacity as well as battery life of nodes. Beginning with the Foschini-Miljanic (FM) algorithm [1], distributed power control for wireless networks has been an active area of research. The problem of distributed power control for packetized traffic was first addressed by Kandukuri and Bambos [2] by proposing “Power-controlled multiple access” (PCMA) family of algorithms, and subsequently investigated in [3], [5], [6]. PCMA algorithm uses dynamic programming approach to balance the power versus delay cost of the transmitter. It introduces three separate regimes: aggressive, soft backoff, hard backoff. In the low-interference regime, transmitter power tries to aggressively match and overcome the interference – up to a certain point, while in the high interference regime, the power is set to zero (hard backoff) and the transmitter ceases to transmit waiting for the low interference periods. Thus, PCMA algorithms usually induce a soft time-division multiple access (iTDMA) allowing links facing high interference to temporarily reduce their power and incur a delay cost instead. On the other hand, in Foschini-Miljanic (FM) algorithm, each link attempts to maintain a target SIR by overcoming the interference presented by the other links. In this project, distributed power control algorithms are investigated in a two-link interference-limited network with asymmetric interference. Such asymmetric interference can occur in any cellular as well as adhoc wireless networks, when the first link receives strong interference from the second link, and the second link receives weak interference from the first link. For example, a downlink user connected to a distant macro-cell can be jammed due to the presence of a closer downlink user connected to a femtocell using the same frequency/time slot (see Fig. 1(a)). In adhoc wireless networks, the random placement of links may also result in asymmetric interference. Since distributed power control algorithms control transmitter power based on the interference seen by the transmitter, they can perform poorly in asymmetric interference-limited networks. If FM algorithm is used in asymmetric interference-limited networks, the weak link uses high power proportional to the interference from the strong link to maintain its target SIR, and its queue starts exploding at very low loads. PCMA algorithm performs better than FM algorithm in asymmetric interference-limited network, but it is unable to induce a iTDMA effect because the strong link can get stuck in aggressive or soft backoff regime when the average load is high and interference seen is low. Distributed power control algorithms have not been investigated in asymmetric interference-limited networks to the best of my knowledge. In this project, distributed power control algorithms [1], [3] are simulated, and their performance is analyzed for two-link asymmetric interference-limited network. Based on the results, some variations are proposed to improve the performance of PCMA algorithms, and these variations are still under investigation.