An Experimental Case for SIMO Random Access in Multi-hop Wireless Networks Ahmed Khattab The Center for Advanced Computer Studies (CACS), University of Louisiana at Lafayette, Lafayette, LA 70504 akhattab@louisiana.edu Abstract—In this paper, we demonstrate that multiple con- current asynchronous and uncoordinated Single-Input Multiple- Output (SIMO) transmissions can successfully take place even though the respective receivers do not explicitly null out interfering signals. Thus motivated, we propose simple modifications to the widely deployed IEEE 802.11 MAC to enable multiple non- spatially-isolated SIMO sender-receiver pairs to share the medium. Namely, we propose to increase the physical carrier sensing thresh- old, disable virtual carrier sensing, and enable message in message packet detection. We use experiments to show that while increasing the peak transmission rate, spatial multiplexing schemes such as those employed by the IEEE 802.11n are highly non-robust to asynchronous and uncoordinated interferers. In contrast, we show that the proposed multi-flow SIMO MAC scheme alleviates the severe unfairness resulting from uncoordinated transmissions in 802.11 multi-hop networks. I. I NTRODUCTION Random access Medium Access Control (MAC) protocols are susceptible to packet losses and unfairness in throughput distribution when the competing senders are not able to coor- dinate their transmissions. Increasing the underlying physical layer rate increases the rate of only MAC contention-winning flows. However, flows which are not able to win medium access still suffer very low throughput, despite their potentially high physical layer rate as was experimentally shown in [1] using commodity IEEE 802.11n hardware. A key reason for the failure of the IEEE 802.11n to provide fairness is that the multi-antenna physical layer is used only to increase the per-link throughput (assuming fading and receiver noise are the only sources of randomness). However, such a physical layer does not counter uncoordinated or hidden senders, but instead relies on the 802.11 MAC to prevent their negative effects. In multi-hop random access networks, wherein senders are not necessarily within range, asynchronous interference from concurrent transmissions is unavoidable since nodes necessarily take uncoordinated transmission actions (e.g., starting time, power, or rate, etc.). In this paper, we consider the additional source of ran- domness at the physical layer resulting from random and unpredictable interference from uncoordinated transmissions, and ask what is the best use of multiple antennas in this case. Unlike related information-theoretic [2]–[4] and MIMO MAC [5]–[8] approaches, we consider the uncoordinated interference scenario wherein a flow is unable to infer the interferers’ chan- nels, rates, power selections, or starting times. The contributions of the paper are as follows. First, we experimentally demonstrate that in the presence of uncoordinated interference, Single-Input Multiple-Output (SIMO) receive diversity provides increased robustness for a wide range of signal to interference ratio (SIR) compared to MIMO spatial multiplexing schemes. Such schemes require significantly high SIR margin at the receiver to attain the promised gains. Thus, SIMO robust transmission is suitable when senders do not apriori know the SIR at the receivers. For example, using the WARP platform [9], we show that a 4 × 4 MIMO flow has to be more than 15 dB above an uncoordinated interferer to attain the 4× throughput gain. Meanwhile, a 1 × 4 SIMO flow’s transmission is almost error-free at -5 dB SIR. Furthermore, we show that for a given cumulative interference, SIMO reliability worsens with fewer uncoordinated interferers. Therefore, allowing more SIMO transmissions to take place at low transmission power is less harmful to the ongoing transmissions than few high-power ones. Second, we show how simple modifications to the IEEE 802.11n protocol can exploit the reliability of SIMO links to alleviate the consequences of uncoordinated transmissions. The main idea is to allow for multiple asynchronous spatially- proximate SIMO transmissions to take place rather than at- tempting to ensure that a single spatially-isolated flow uses MIMO to increase its rate (as the case with 802.11n). We show that this can be achieved by suitably increasing the physical carrier sensing threshold and disabling virtual carrier sensing. 1 Consequently, a sender can initiate a new transmission – even if other nearby transmissions are currently taking place – as long as the cumulative interference in its vicinity implies a suffi- cient interference margin for an additional SIMO transmission. Meanwhile, a receiver must be able to lock on to a new arriving packet after receiving the preamble of an unintended packet. Thus, enabling the Message-In-Message (MIM) [10] 802.11 feature is mandatory for our multi-flow MAC approach. Our results show that our SIMO MAC alleviates the unfairness of legacy MIMO MAC in problematic topologies. The paper is organized as follows. We motivate the SIMO case in Section II. In Section III, we experimentally demon- strate SIMO robustness to uncoordinated interference. Then, we present simple modifications to 802.11 to exploit SIMO flows to alleviate unfairness in Section IV. We discuss related work in Section V and conclude in Section VI. II. SYSTEM MODEL AND MOTIVATION In this section, we define the system model and the uncoor- dinated transmissions’ problems in random access networks. A. System Model We consider a multi-hop ad hoc network wherein individual sender-receiver pairs may not be mutually within transmission range of each other. Each node is equipped with a half-duplex transceiver with N> 1 antennas tuned to the same frequency. 1 Virtual carrier sensing prohibits a node from transmitting after an overheard 802.11 packet header is decoded until after the packet’s duration field indicates that the transmission will be completed. This paper was presented as part of the Mini-Conference at IEEE INFOCOM 2010 This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2010 proceedings 978-1-4244-5837-0/10/$26.00 ©2010 IEEE