This letter addresses multisensor data fusion under the Gaussian noise. Under the Gauss-Markov model assumptions, data fusion based on maximum likelihood estimation (MLE) is the minimum variance unbiased estimator. Nonetheless, we propose a linear fusion algorithm based on the random matrix theory, which yields a biased estimator. The proposed estimator has a lower mean squared error (MSE) than the MLE fusion method when the dimensionality of signal is larger than the number of sensors, which is the typical use case in modern fusion systems. The fusion coefficients are directly solved in the proposed method without iteration, and this method can be considered as an approximate implementation of the linear minimum MSE (LMMSE) estimator. Numerical simulations demonstrate the performance gain of the proposed fusion method.

Existing work on pursuit-evasion problems typically either assumes stationary or heuristic behavior of one side and examines countermeasures of the other, or assumes both sides to be strategic which leads to a game theoretical framework. Results from the former often lack robustness against changes in the adversarial behavior, while those from the second category, typically as equilibrium solution concepts, may be difficult to justify: either due to the implied knowledge of other players' actions/beliefs and knowledge of their knowledge, or due to a lack of efficient dynamics to achieve such equilibria. In this paper, we take a different approach by assuming an intelligent pursuer/evader that is adaptive to the information available to it and is capable of learning over time with performance guarantee. Within this context we investigate two cases. In the first case we assume either the evader or the pursuer is aware of the type of learning algorithm used by the opponent, while in the second case neither side has such information and thus must try to learn. We show that the optimal policies in the first case have a greedy nature. This result is then used to assess the performance of the learning algorithms that both sides employ in the second case, which is shown to be mutually optimal and there is no loss for either side compared to the case when it knows perfectly the adaptive pattern used by the adversary and responses optimally. We further extend our model to study the application of jamming defense.

Sniffer channel assignment (SCA) is a fundamental building block for wireless data capture, which is essential for traffic monitoring and network forensics. Most of the existing SCA approaches for cognitive radio networks (CRNs) adopt optimization-based methods and rely on the prior knowledge of the secondary user (SU) activities. To relax this constraint, learning-based methods have been recently developed; however, there is still insufficient theoretical understanding within the learning framework for SCA. In this paper, we aim to maximize the total amount of the captured SU traffic, and we formulate the SCA problem as a nonstochastic/adversarial multiarmed bandit problem. Moreover, the inherent error in wireless capturing, i.e., imperfect monitoring, is considered in our model. We propose two online learning algorithms for the SCA scenarios with and without channel switching costs, respectively, and their regret performances are proved uniformly sublinear in time and polynomial in the number of channels. The numerical evaluation shows, in addition to their robust regret performances, the proposed algorithms greatly outperform the existing SCA approaches in the amount of effectively captured SU traffic.

Existing work on pursuit-evasion problems typically either assumes stationary or heuristic behavior of one side and examines countermeasures of the other, or assumes both sides to be strategic which leads to a game theoretical framework. Results from the former may lack robustness against changes in the adversarial behavior, while those from the latter are often difficult to justify due to the implied full information (either as realizations or as distributions) and rationality, both of which may be limited in practice. In this paper, we take a different approach by assuming an intelligent pursuer/evader that is adaptive to the information available to it and is capable of learning over time with performance guarantee. Within this context we investigate two cases. In the first case we assume either the evader or the pursuer is aware of the type of learning algorithm used by the opponent, while in the second case neither side has such information and thus must try to learn. We show that the optimal policies in the first case have a greedy nature, hiding/seeking in the location that the opponent is the least/most likely to appear. This result is then used to assess the performance of the learning algorithms that both sides employ in the second case, which is shown to be mutually optimal and there is no loss for either side compared to the case when it completely knows the adaptive pattern used by the adversary and responses optimally.

We revisit the jamming defense problem in a multi-channel wireless system, using a general formulation of online learning against an adversary via repeated game-playing. We provide the explicit form of the worst-case optimal channel-hopping strategy of a legitimate user in a multi-stage interaction with a resource-replenishing jamming attacker. Interestingly, we show that the worst imaginary enemy can be given as an adversary who behaves in an i.i.d. manner in this multi-stage interaction, and the optimal strategy of the user is determined by the induced random walk of the adversarial behavior. In addition to the jamming defense, our framework is also applicable to other competitive game problems with finite action spaces.

We consider the problem of monitoring soil moisture evolution using a wireless network of in-situ underground sensors. To reduce cost and prolong lifetime, it is highly desirable to rely on fewer measurements and estimate with higher accuracy the original signal (the temporal evolution of soil moisture). In this article, we explore the use of results from the theory of sparse sampling, including Compressive Sensing (CS) and Matrix Completion (MC), in this application context. We first consider the problem of reconstructing the soil moisture process at a single location using CS. Our physical constraint leads to very sparse measurement matrices, which makes finding a suitable representation basis very challenging: it needs to make the underlying signal sufficiently sparse while at the same time being sufficiently incoherent with the measurement matrix, two common preconditions for CS techniques to work well. We construct a representation basis by exploiting unique features of soil moisture evolution and show that this basis attains a very good tradeoff between its ability to sparsify the signal and its incoherence with measurement matrices that are consistent with our physical constraints. We next consider the problem of jointly reconstructing soil moisture processes at multiple locations, assuming sparse measurements can be taken at each location. We show that the spatial soil moisture process enjoys a low-rank property, a priority for MC. Accordingly, we introduce a spatiotemporal measurement matrix and apply the MC framework to reconstruct the soil moisture field. Extensive numerical evaluation is performed on both real, high-resolution soil moisture data and simulated data and through comparison with a closed-loop scheduling approach. Our results demonstrate that, for a single location, a uniform measurement scheduling followed by CS recovery results in a very nice tradeoff between estimation accuracy, sampling rate, flexibility, and feasibility in implementation. When multiple locations are available, our results show that joint reconstruction using MC in general produces better estimation accuracy than using a single location alone, but it requires the use of independent and random measurement schedules across locations. We also show that these sparse sampling techniques can be augmented so as to be robust against sporadic data outliers/corruption caused by, for example, intermittent sensor faults.

We consider the optimal transmission power control of a single wireless node with stochastic energy harvesting and an infinite/saturated queue with the objective of maximizing a certain reward function, e.g., the total data rate. We develop simple control policies that achieve near optimal performance in the finite-horizon case with finite energy storage. The same policies are shown to be asymptotically optimal in the infinite horizon case for sufficiently large energy storage. Such policies are typically difficult to directly obtain using a Markov Decision Process (MDP) formulation or through a dynamic programming framework due to the computational complexity. We relate our results to those obtained in the unsaturated regime, and highlight a type of threshold-based policies that is universally optimal.

We observe that in a multichannel wireless system, an opportunistic channel/spectrum access scheme that solely focuses on channel quality sensing measured by received SNR may induce users to use channels that, while providing better signals, are more congested. Ultimately the notion of channel quality should include both the signal quality and the level of congestion, and a good multichannel access scheme should take both into account in deciding which channel to use and when. Motivated by this, we focus on the congestion aspect and examine what type of dynamic channel switching schemes may result in the best system throughput performance. Specifically, we derive the stability region of a multiuser multichannel WLAN system and determine the throughput optimal channel switching scheme within a certain class of schemes. We also empirically examine the impact of considering congestion in addition to signal quality in making channel selection decisions.

In this paper, we investigate the connectivity for large-scale clustered wireless sensor and ad hoc networks. We study the effect of mobility on the critical transmission range for asymptotic connectivity ink-hop clustered networks and compare to existing results on nonclustered stationary networks. By introducingk-hop clustering, any packet from a cluster member can reach a cluster head withinkhops, and thus the transmission delay is bounded as Θ(1) for any finitek. We first characterize the critical transmission range for connectivity in mobilek-hop clustered networks where all nodes move under either the random walk mobility model with nontrivial velocity or the i.i.d. mobility model. By the term nontrivial velocity, we mean that the velocity of a nodevis ω(r(n)), wherer(n) is the transmission range of the node. We then compare with the critical transmission range for stationaryk-hop clustered networks. In addition, the critical number of neighbors is studied in a parallel manner for both stationary and mobile networks. We also study the transmission power versus delay tradeoff and the average energy consumption per flow among different types of networks. We show that random walk mobility with nontrivial velocities increases connectivity ink-hop clustered networks, and thus significantly decreases the energy consumption and improves the power-delay tradeoff. The decrease of energy consumption per flow is shown to be Θ([(logn)/(nd)]) in clustered networks. These results provide insights on network design and fundamental guidelines on building a large-scale wireless network.

We observe that in a multi-channel system, an opportunistic channel access scheme that solely focuses on channel quality sensing measured by received SNR may induce users to use channels that, while providing better signals, are more congested. Ultimately the notion of channel quality should include both the signal quality and the level of congestion, and a good multi-channel access scheme should take both into account in deciding which channel to use and when. Motivated by this, we focus on the congestion aspect and examine what type of dynamic channel switching schemes may result in the best system throughput performance. Specifically we derive the stability region of a multi-user multi-channel WLAN system and determine the throughput optimal channel switching scheme within a certain class of schemes.

In this paper we investigate the connectivity for large-scale clustered wireless sensor and ad hoc networks. We study the effect of mobility on the critical transmission range for asymptotic connectivity in k-hop clustered networks, and compare to existing results on non-clustered stationary networks. By introducing k-hop clustering, any packet from a cluster member can reach a cluster head within k hops, and thus the transmission delay is bounded as Θ(1) for any finite k. We first characterize the critical transmission range for connectivity in mobile k-hop clustered networks where all nodes move under either the random walk mobility model with non-trivial velocity or the i.i.d. mobility model. By the term non-trivial velocity, we mean that the velocity of nodes v is Θ(1). We then compare with the critical transmission range for stationary k-hop clustered networks. We also study the transmission power versus delay trade-off and the average energy consumption per flow among different types of networks. We show that random walk mobility with non-trivial velocity increases connectivity in k-hop clustered networks, and thus significantly decreases the energy consumption and improves the power-delay trade-off. The decrease of energy consumption per flow is shown to be Θ(logn/nd}) in clustered networks. These results provide insights on network design and fundamental guidelines on building a large-scale wireless network.