Research Article
Transport Capacity of Opportunistic Spectrum Access (OSA) MANETs
@INPROCEEDINGS{10.1109/CROWNCOM.2007.4549765, author={Cesar Santivanez}, title={Transport Capacity of Opportunistic Spectrum Access (OSA) MANETs}, proceedings={2nd International ICST Conference on Cognitive Radio Oriented Wireless Networks and Communications}, publisher={IEEE}, proceedings_a={CROWNCOM}, year={2008}, month={6}, keywords={Bandwidth Computer networks Frequency Government Interference Mobile ad hoc networks Network topology Protection Throughput}, doi={10.1109/CROWNCOM.2007.4549765} }- Cesar Santivanez
Year: 2008
Transport Capacity of Opportunistic Spectrum Access (OSA) MANETs
CROWNCOM
IEEE
DOI: 10.1109/CROWNCOM.2007.4549765
Abstract
In this paper, we study the Transport Capacity (TC) of a Mobile Ad hoc NETwork (MANET) of secondary nodes without a predefined bandwidth assignment. These secondary users' access of the spectrum is governed by Opportunistic Spectrum Access (OSA) rules, i.e., they can use portions of the spectrum as long as their transmissions do not interfere with the spectrum's primary (protected) users. We present an analytical framework to compute the TC (bits-meter per second) of OSA MANETs. Our results show that an OSA MANET exhibits two operating regions, determined by the ratio of the secondary nodes' transmission range (ls) over the primary nodes' (lp). When ls ≪ lp, the OSA MANET is in the interference limited region, and its TC behaves exactly as a legacy MANET, that is, its TC decays slowly with respect to ls: TC ÿ k A W/ls = ÿ (ls-1), where A is the network area, and W is the usable spectrum. However, when ls ≫ lp, the OSA MANET is in the policy limited region, where TC decays faster than before w.r.t. ls, that is, TC ÿ kÿ A W/ls (lp/ls)ÿ-2 = ÿ (ls-(ÿ-1)) if density is unbounded, and TC ÿ kÿ A W/ls (lp/ls)ÿ = ÿ (ls-(ÿ+1)) if density is bounded, where ÿ ≫ 2 is the pathloss exponent. These results are of great importance to understand the behavior of OSA-enabled MANETs, especially when designing a network (e.g., number of nodes, density, needed to cover an area) or developing self-optimizing algorithms (e.g., topology control).
