The integration of wireless sensors with mobile phones is gaining momentum as an enabling platform for numerous emerging applications. These mobile systems face dynamic environments where both application requirements and ambient wireless conditions change frequently. Despite the existence of many MAC protocols, none can provide optimal characteristics along multiple dimensions, especially when the conditions are frequently changing. Instead of pursuing a one-MAC-fit-all approach we present the Self-Adapting MAC Layer (SAML) that dynamically selects and switches MAC protocols to gain the desired characteristics in response to changes in ambient conditions and application requirements. SAML comprises (1) a Reconfigurable MAC Architecture (RMA) that can switch to different MAC protocols at run time and (2) a learning-based MAC Selection Engine that selects the protocol most suitable for the current condition and requirements. To the application SAML appears as a traditional MAC layer and realizes its benefits through a simple API for the mobile applications. We have implemented SAML in TinyOS 2.x and built three prototypes containing up to five MACs. We evaluate the system in controlled tests and real-world environments using a new gateway device that integrates a 802.15.4 radio with Android phones. Our experimental results show that SAML can effectively adapt MAC layer behavior to meet varying application requirements in dynamic environments through judicious selection and efficient switching of MAC protocols.

In multiuser multiple-input-multiple-output (MIMO) networks, receivers decode multiple concurrent signals using successive interference cancellation (SIC). With SIC, a weak target signal can be deciphered in the presence of stronger interfering signals. However, this is only feasible if each strong interfering signal satisfies a signal-to-noise-plus-interference ratio (SINR) requirement. This necessitates the appropriate selection of a subset of links that can be concurrently active in each receiver's neighborhood; in other words, a subtopology consisting of links that can be simultaneously active in the network is to be formed. If the selected subtopologies are of small size, the delay between the transmission opportunities on a link increases. Thus, care should be taken to form a limited number of subtopologies. We find that the problem of constructing the minimum number of subtopologies such that SIC decoding is successful with a desired probability threshold is NP-hard. Given this, we propose MUSIC, a framework that greedily forms and activates subtopologies in a way that favors successful SIC decoding with a high probability. MUSIC also ensures that the number of selected subtopologies is kept small. We provide both a centralized and a distributed version of our framework. We prove that our centralized version approximates the optimal solution for the considered problem. We also perform extensive simulations to demonstrate that: 1) MUSIC forms a small number of subtopologies that enable efficient SIC operations; the number of subtopologies formed is at most 17% larger than the optimum number of topologies, discovered through exhaustive search (in small networks); 2) MUSIC outperforms approaches that simply consider the number of antennas as a measure for determining the links that can be simultaneously active. Specifically, MUSIC provides throughput improvements of up to four times, as compared to such an approach, in various topological settings. The improvements can be directly attributable to a significantly higher probability of correct SIC based decoding with MUSIC.

Network coding has been proposed as a technique that can potentially increase the transport capacity of a wireless network via mixing data packets at intermediate routers. However, most previous studies either assume a fixed transmission rate or do not consider the impact of using diverse rates on the network coding gain. Since in many cases, network coding implicitly relies on overhearing, the choice of the transmission rate has a big impact on the achievable gains. The use of higher rates works in favor of increasing the native throughput. However, it may in many cases work against effective overhearing. In other words, there is a tension between the achievable network coding gain and the inherent rate gain possible on a link. In this paper, our goal is to drive the network toward achieving the best tradeoff between these two contradictory effects. We design a distributed framework that: facilitates the choice of the best rate on each link while considering the need for overhearing; and dictates the choice of which decoding recipient will acknowledge the reception of an encoded packet. We demonstrate that both of these features contribute significantly toward gains in throughput. We extensively simulate our framework in a variety of topological settings. We also fully implement it on real hardware and demonstrate its applicability and performance gains via proof-of-concept experiments on our wireless testbed. We show that our framework yields throughput gains of up to 390% as compared to what is achieved in a rate-unaware network coding framework.

Many next generation applications (such as video flows) are likely to have associated minimum data rate requirements in order to ensure satisfactory quality as perceived by end-users. In this paper, we develop a framework to address the problem of maximizing the aggregate utility of traffic flows in a multi-hop wireless network, with constraints imposed both due to self-interference and minimum rate requirements. The parameters that are tuned in order to maximize the utility are (i) transmission powers of individual nodes and (ii) the channels assigned to the different communication links. Our framework is based on using a cross-decomposition technique that takes both inter-flow interference and self-interference into account. The output of our framework is a schedule that dictates what links are to be activated in each slot and the parameters associated with each of those links. If the minimum rate constraint cannot be satisfied for all of the flows, the framework intelligently rejects a sub-set of the flows and recomputes a schedule for the remaining flows. We also design an admission control module that determines if new flows can be admitted without violating the rate requirements of the existing flows in the network. We provide numerical results to demonstrate the efficacy of our framework.

With the rapid permeation of smartphones and wireless sensors in our society, smartphones are poised to become personal hubs connecting wireless sensors with users and the Internet. Due to frequent changes to applications and network conditions, wireless networks connecting a personal hub and wireless sensors must meet time-varying QoS requirements at minimal energy cost. This paper presents the architecture of a novel Protocol Service System (PSS) towards the vision of protocol service over the air. In contrast to traditional wireless networks where a single MAC protocol is statically selected a priori, PSS switches among multiple MAC protocols at run time to dynamically optimize power efficiency subject and meet current QoS requirements. To meet the memory constraint on wireless sensors PSS employs a component-based reconfigurable MAC architecture to support multiple MAC protocols at significantly reduced memory footprint through component sharing. The feasibility of PSS has been demonstrated through a proof-of-concept implementation of the PSS architecture and the development of a personal hub prototype running the Android OS.

In Multi-User MIMO networks, receivers decode multiple concurrent signals using Successive Interference Cancellation (SIC). With SIC a weak target signal can be deciphered in the presence of stronger interfering signals. However, this is only feasible if each strong interfering signal satisfies a signal-to-noise-plus-interference ratio (SINR) requirement. This necessitates the appropriate selection of a subset of links that can be concurrently active in each receiver's neighborhood; in other words, a sub-topology consisting of links that can be simultaneously active in the network is to be formed. If the selected sub-topologies are of small size, the delay between the transmission opportunities on a link increases. Thus, care should be taken to form a limited number of sub-topologies. We find that the problem of constructing the minimum number of sub-topologies such that SIC decoding is successful with a desired probability threshold, is NP-hard. Given this, we propose MUSIC, a framework that greedily forms and activates sub-topologies, in a way that favors successful SIC decoding with a high probability. MUSIC also ensures that the number of selected sub-topologies is kept small. We provide both a centralized and a distributed version of our framework. We prove that our centralized version approximates the optimal solution for the considered problem. We also perform extensive simulations to demonstrate that (i) MUSIC forms a small number of sub-topologies that enable efficient SIC operations; the number of sub-topologies formed is at most 17% larger than the optimum number of topologies, discovered through exhaustive search (in small networks). (ii) MUSIC outperforms approaches that simply consider the number of antennas as a measure for determining the links that can be simultaneously active. Specifically, MUSIC provides throughput improvements of up to 4 times, as compared to such an approach, in various topological settings. The improvements can be directly attributable to a significantly higher probability of correct SIC based decoding with MUSIC.

Network coding has been proposed as a technique that can potentially increase the transport capacity of a wireless network via processing and mixing of data packets at intermediate routers. However, most previous studies either assume a fixed transmission rate or do not consider the impact of using diverse rates on the network coding gain. Since in many cases, network coding implicitly relies on overhearing, the choice of the transmission rate has a big impact on the achievable gains. The use of higher rates works in favor of increasing the native throughput; however, it may in many cases work against effective overhearing. In other words, there is a tension between the achievable network coding gain and the inherent rate gain possible on a link. In this paper our goal is to drive the network towards achieving the best trade-off between these two contradictory effects. Towards this, we design a distributed framework that (a) facilitates the choice of the best rate on each link while considering the need for overhearing and (b) dictates the choice of which decoding recipient will acknowledge the reception of an encoded packet. We demonstrate that both of these features contribute significantly towards gains in throughput. We extensively simulate our framework in a variety of topological settings. We also fully implement it on real hardware and demonstrate its applicability and performance gains via proof-of-concept experiments on our wireless testbed. We show that our framework yields throughput gains of up to 390% as compared to what is achieved in a rate-unaware network coding framework.

The unlicensed 60 GHz band brings the promise of multi-gigabit data rates to support new applications such as high definition video over wireless links. Signal propagation in the 60 GHz band significantly differs from that in the traditionally used 2.4 and 5 GHz bands. The propagation and penetration losses in the 60 GHz band are much higher. Furthermore, the signals are often reflected in indoor settings. Previous physical layer studies show that the use of directional antennas can significantly help in coping with these effects. In this paper, we address the problem of neighbor discovery in the 60 GHz band. We account for not only discovery via direct line-of-sight paths, but also via reflected beams. To the best of our knowledge, none of the previous efforts on higher layer protocols for use with directional antennas account for reflections. We consider two approaches for neighbor discovery (a) direct discovery where each node explicitly discovers its neighbors and, (b) gossip-based discovery where nodes exchange information with regards to their already discovered neighbors. We develop analytical models to capture the performance of the two approaches and validate the models via simulations in indoor settings with obstacles that reflect the transmitted signals. As one might expect, the gossip based discovery incurs a lower neighbor-discovery latency than direct discovery. We examine the impact of system parameters such as varying beamwidth and node density. Our study provides insights on the right choice of system parameters for efficient neighbor discovery in the 60 GHz regime.

Many next generation applications (such as video flows) are likely to have associated minimum data rate requirements to ensure satisfactory quality as perceived by end-users. While there have been prior approaches on supporting quality-of-service (QoS) in mesh networks, they have largely ignored the issues that arise due to self-interference, the interference between different link layer transmissions of a single flow along a multi-hop path. In this paper, we develop a framework to address the problem of maximizing the aggregate utility of traffic flows in wireless mesh networks, with constraints imposed both due to self-interference and minimum rate requirements. The output of our framework is a schedule that dictates which links are to be activated simultaneously, and provides specifications of the resources associated with each of those links. Utilizing the proposed framework as a basis, we build an admission control module that intelligently manages the resources among the flows in the network and admits as many new flows as possible without violating the QoS of the existing flows. We provide numerical results to demonstrate the efficacy of our framework.

The importance of spatial reuse in wireless ad hoc networks has been long recognized as a key to improving the network capacity. In this paper, we show that 1) in the case that the achievable channel rate follows the Shannon capacity, spatial reuse depends only on the ratio of the transmit power to the carrier sense threshold and 2) in the case that only a set of discrete data rates are available, as a control knob for sustaining achievable data rates, tuning the transmit power provides more sophisticated rate control over tuning the carrier sense threshold, provided that there is a sufficient number of power levels available. Based on the findings, we then propose a decentralized power and rate control algorithm to enable each node to adjust, based on its signal interference level, its transmit power and data rate. The transmit power is so determined that the transmitter can sustain a high data rate while keeping the adverse interference effect on the other neighboring concurrent transmissions minimal. Simulation results have shown that compared to existing carrier sense threshold tuning algorithms, the proposed power and rate control algorithm yields higher network capacity.

The importance of spatial reuse in wireless ad-hoc networks has been long recognized as a key to improving the network capacity. One can increase the level of spatial reuse by either reducing the transmit power or increasing the carrier sense threshold (thereby reducing the carrier sense range). On the other hand, as the transmit power decreases or the carrier sense threshold increases, the SINR decreases as a result of the smaller received signal or the increased interference level. Consequently, the data rate sustained by each transmission may decrease. This leads naturally to the following questions:(1)How can the trade-off between the increased level of spatial reuse and the decreased data rate each node can sustain be quantified? In other words,is there an optimal range of transmit power/carrier sense threshold in which the network capacity is maximized? (2)What is the relation between the transmit power and the carrier sense threshold.In this paper, we study both problems, and show that (i)in the case that the achievable channel rate follows the Shannon capacity, spatial reuse depends only on the ratio of the transmit power to the carrier sense threshold; and (ii) in the case that only a set of discrete data rates are available, tuning the transmit power offers several advantages that tuning the carrier sense threshold cannot, provided that there is a sufficient number of power levels available. Based on the findings, we then propose a decentralized power and rate control algorithm to enable each node to adjust, based on its signal interference level, its transmit power and data rate. The transmit power is so determined that the transmitter can sustain a high data rate, while keeping the adverse interference effect on the other neighboring concurrent transmissions minimal. Simulation results have shown that, as compared to existing carrier sense threshold tuning algorithms, the proposed power and rate control algorithm yields higher network capacity.