John Heidemann

Evaluating the Importance of Concurrent Packet Communication in Wireless Networks

TitleEvaluating the Importance of Concurrent Packet Communication in Wireless Networks
Publication TypeTechnical Report
Year of Publication2007
AuthorsD. Son, B. Krishnamachari, and J. Heidemann
Date Publishedapr

Nearly all wireless media-access (MAC) protocols are designed today with the very conservative assumption that concurrent transmissions should be prevented, because sender-receiver pairs within radio range sending on the same channel will corrupt each other's communication. While recent work has suggested that channel capture effects can be significant in reality, this paper presents the first systematic study to quantify the impact of these effects on the ability to have concurrent communications among two sender-receiver pairs that are within range of each other. We first derive a simple decision rule to determine when such concurrent communication is possible while minimizing transmission power. Through a comprehensive set of realistic simulations, we then systematically quantify the feasibility of concurrent communication with and without transmission power control as many radio and environmental parameters vary, including node position, mean and variance of path loss, signal-to-interference-plus-noise-ratio threshold ($SINR_þeta$) for packet reception, granularity and range of transmission power control. Our simulations show that often, 40–75% of the time depending primarily on distance and location, two pairs of nodes can communicate concurrently without loss even if both transmitters are within the radio sensing range of both receivers. We can observe large CTXable region with fixed transmit power, but dynamic power control significantly improves concurrent communications. Finally, at least one transmitter can almost always capture the channel in the event of concurrent transmissions, so the cost of failed attempts to CTX are minimal. We validate our simulations with testbed experiments using MicaZ motes, confirming that concurrent communication is possible to a very significant extent in real systems. These results suggest that CSMA with RTS/CTS is overly conservative and there are often gains to be realized by abandoning it.