Search

We consider multi--hop wireless networks carrying unicast flows for multiple users. Each flow has a specified delay deadline, and the lossy wireless links are modelled as binary symmetric channels (BSCs). Since transmission time, also called airtime, on the links is shared amongst flows, increasing the airtime for one flow comes at the cost of reducing the airtime available to other flows sharing the same link. We derive the joint allocation of flow airtimes and coding rates that achieves the proportionally fair throughput allocation. This utility optimisation problem is non--convex, and one of the technical contributions of this paper is to show that the proportional fair utility optimisation can nevertheless be decomposed into a sequence of convex optimisation problems. The solution to this sequence of convex problems is the unique solution to the original non--convex optimisation. Surprisingly, this solution can be written in an explicit form that yields considerable insight into the nature of the proportional fair joint airtime/coding rate allocation. To our knowledge, this is the first time that the utility fair joint allocation of airtime/coding rate has been analysed, and also, one of the first times that utility fairness with delay deadlines has been considered., Comment: Submitted to IEEE/ACM Transactions on Networking (December 2011)

We define a class of multi--hop erasure networks that approximates a wireless multi--hop network. The network carries unicast flows for multiple users, and each information packet within a flow is required to be decoded at the flow destination within a specified delay deadline. The allocation of coding rates amongst flows/users is constrained by network capacity. We propose a proportional fair transmission scheme that maximises the sum utility of flow throughputs. This is achieved by {\em jointly optimising the packet coding rates and the allocation of bits of coded packets across transmission slots.}, Comment: Submitted to the Forty-Ninth Annual Allerton Conference on Communication, Control, and Computing, Monticello, Illinois, USA

We consider multi--hop networks comprising Binary Symmetric Channels ($\mathsf{BSC}$s). The network carries unicast flows for multiple users. The utility of the network is the sum of the utilities of the flows, where the utility of each flow is a concave function of its throughput. Given that the network capacity is shared by the flows, there is a contention for network resources like coding rate (at the physical layer), scheduling time (at the MAC layer), etc., among the flows. We propose a proportional fair transmission scheme that maximises the sum utility of flow throughputs subject to the rate and the scheduling constraints. This is achieved by {\em jointly optimising the packet coding rates of all the flows through the network}., Comment: Submitted to Forty-Ninth Annual Allerton Conference on Communication, Control, and Computing, Monticello, IL, USA

We study a problem of distributed detection of a stationary point event in a large extent wireless sensor network ($\wsn$), where the event influences the observations of the sensors only in the vicinity of where it occurs. An event occurs at an unknown time and at a random location in the coverage region (or region of interest ($\ROI$)) of the $\wsn$. We consider a general sensing model in which the effect of the event at a sensor node depends on the distance between the event and the sensor node; in particular, in the Boolean sensing model, all sensors in a disk of a given radius around the event are equally affected. Following the prior work reported in \cite{nikiforov95change_isolation}, \cite{nikiforov03lower-bound-for-det-isolation}, \cite{tartakovsky08multi-decision}, {\em the problem is formulated as that of detecting the event and locating it to a subregion of the $\ROI$ as early as possible under the constraints that the average run length to false alarm ($\tfa$) is bounded below by $\gamma$, and the probability of false isolation ($\pfi$) is bounded above by $\alpha$}, where $\gamma$ and $\alpha$ are target performance requirements. In this setting, we propose distributed procedures for event detection and isolation (namely $\mx$, $\all$, and $\hall$), based on the local fusion of $\CUSUM$s at the sensors. For these procedures, we obtain bounds on the maximum mean detection/isolation delay ($\add$), and on $\tfa$ and $\pfi$, and thus provide an upper bound on $\add$ as $\min\{\gamma,1/\alpha\} \to \infty$. For the Boolean sensing model, we show that an asymptotic upper bound on the maximum mean detection/isolation delay of our distributed procedure scales with $\gamma$ and $\alpha$ in the same way as the asymptotically optimal centralised procedure \cite{nikiforov03lower-bound-for-det-isolation}., Comment: Submitted to IEEE Transactions on Signal Processing, Mar. 10, 2011. Revised on Jul. 17, 2011. A part of this work was presented in Forty-Seventh Annual Allerton Conference on Communication, Control, and Computing, Monticello, IL, USA, Sep. - Oct. 2009

We consider the problem of quickest event detection with sleep-wake scheduling in small extent wireless sensor networks in which, at each time slot, each sensor node in the awake state observes a sample and communicates the information to the fusion centre. The sensor nodes in the sleep state do not sample or communicate any information to the fusion centre (FC), thereby conserving energy. At each time slot, the FC, after having received the samples from the sensor nodes in the wake state, makes a decision to stop (and thus declare that the event has occurred) or to continue observing. If it decides to continue, the FC also makes the decision of choosing the number of sensor nodes to be in the wake state in the next time slot. We consider three alternative approaches to the problem of choosing the number of sensor nodes to be in the wake state in time slot k+1, based on the information available at time slot k, namely, 1. optimal control of M_{k+1}, the number of sensor nodes to be in the awake state in time slot k+1, 2. optimal control of q_{k+1}, the probability of a sensor node to be in the awake state in time slot k+1, and 3. optimal probability q that a sensor node is in the awake state in any time slot. In each case, we formulate the problem as a sequential decision process. We show that a sufficient statistic for the decision at time k is the a posteriori probability of change Pi_k. Also, we show that the optimal stopping rule is a threshold rule on Pi_k. The optimal policy for M_{k+1} can keep very few sensors wake during the prechange phase and then quickly increase the number of sensors in the wake state when a change is "suspected". Among the three sleep-wake algorithms described, we observe that the total cost is minimum for the optimum control of M_{k+1} and is maximum for the optimum control on q., Comment: Submitted to IEEE/ACM Transactions on Networking, Feb. 01, 2011. This is an expanded version of a paper that was presented in IEEE INFOCOM 2008

We consider a small extent sensor network for event detection, in which nodes take samples periodically and then contend over a {\em random access network} to transmit their measurement packets to the fusion center. We consider two procedures at the fusion center to process the measurements. The Bayesian setting is assumed; i.e., the fusion center has a prior distribution on the change time. In the first procedure, the decision algorithm at the fusion center is \emph{network-oblivious} and makes a decision only when a complete vector of measurements taken at a sampling instant is available. In the second procedure, the decision algorithm at the fusion center is \emph{network-aware} and processes measurements as they arrive, but in a time causal order. In this case, the decision statistic depends on the network delays as well, whereas in the network-oblivious case, the decision statistic does not depend on the network delays. This yields a Bayesian change detection problem with a tradeoff between the random network delay and the decision delay; a higher sampling rate reduces the decision delay but increases the random access delay. Under periodic sampling, in the network--oblivious case, the structure of the optimal stopping rule is the same as that without the network, and the optimal change detection delay decouples into the network delay and the optimal decision delay without the network. In the network--aware case, the optimal stopping problem is analysed as a partially observable Markov decision process, in which the states of the queues and delays in the network need to be maintained. A sufficient statistic for decision is found to be the network-state and the posterior probability of change having occurred given the measurements received and the state of the network. The optimal regimes are studied using simulation., Comment: To appear in ACM Transactions on Sensor Networks. A part of this work was presented in IEEE SECON 2006, and Allerton 2010