15 September 2016 in IEEE Transactions on Wireless Communications
Millimeter wave (mmWave) systems are emerging as an essential technology for enabling extremely high data rate wireless communications. The main limiting factors of mmWave systems are blockage (high penetration loss) and deafness (misalignment between the beams of the transmitter and receiver). To alleviate these problems, it is imperative to incorporate efficient association and relaying between terminals and access points. Unfortunately, the existing association techniques are designed for the traditional interference-limited networks, and thus are highly suboptimal for mmWave communications due to narrow-beam operations and the resulting non-negligible interference-free behavior. This paper introduces a distributed approach that solves the joint association and relaying problem in mmWave networks considering the load balancing at access points. The problem is posed as a novel stochastic optimization problem, which is solved by distributed auction algorithms where the clients and relays act asynchronously to achieve optimal client-relay-access point association. It is shown that the algorithms provably converge to a solution that maximizes the aggregate logarithmic utility within a desired bound. Numerical results allow quantification of the performance enhancements introduced by the relays, and the substantial improvements of the network throughput and fairness among the clients by the proposed association method as compared to standard approaches. It is concluded that mmWave communications with proper association and relaying mechanisms can support extremely high data rates, connection reliability, and fairness among the clients.
9 September 2022 in IEEE Internet of Things Journal
Inference carried out on pretrained deep neural networks (DNNs) is particularly effective as it does not require retraining and entails no loss in accuracy. Unfortunately, resource-constrained devices such as those in the Internet of Things may need to offload the related computation to more powerful servers, particularly, at the network edge. However, edge servers have limited resources compared to those in the cloud; therefore, inference offloading generally requires dividing the original DNN into different pieces that are then assigned to multiple edge servers. Related approaches in the state-of-the-art either make strong assumptions on the system model or fail to provide strict performance guarantees. This article specifically addresses these limitations by applying distributed assignment to DNN inference at the edge. In particular, it devises a detailed model of DNN-based inference, suitable for realistic scenarios involving edge computing. Optimal inference offloading with load balancing is also defined as a multiple assignment problem that maximizes proportional fairness. Moreover, a distributed algorithm for DNN inference offloading is introduced to solve such a problem in polynomial time with strong optimality guarantees. Finally, extensive simulations employing different data sets and DNN architectures establish that the proposed solution significantly improves upon the state-of-the-art in terms of inference time (1.14 to 2.62 times faster), load balance (with Jain’s fairness index of 0.9), and convergence (one order of magnitude less iterations).
27 November 2018 in IEEE Transactions on Wireless Communications
The underutilized millimeter-wave (mmWave) band is a promising candidate to enable extremely high data rate communications in future wireless networks. However, special characteristics of mmWave systems such as high vulnerability to obstacles (due to high penetration loss) and to mobility (due to directional communications) demand a careful design of the association between clients and access points (APs). This challenge can be addressed by distributed association techniques that gracefully adapt to wireless channel variations and client mobilities. We formulated the association problem as a mixedinteger optimization aiming to maximize the network throughput with proportional fairness guarantees. This optimization problem is solved by first a distributed dual decomposition algorithm, and then by a novel distributed auction algorithm where the clients act asynchronously to achieve near-to-optimal association between clients and APs. The latter algorithm has a faster convergence with a negligible drop in the resulting network throughput. A distinguishing novel feature of the proposed algorithms is that the resulting optimal association does not have to be re-computed every time the network changes (e.g., due to mobility). Instead, the algorithms continuously adapt to the network variations and are thus very efficient. We discuss the implementation of the proposed algorithms on top of existing communication standards. Numerical analysis verifies the ability of the proposed algorithms to optimize the association and to maintain optimality in time-variant environments of mmWave networks.
13 September 2016 in IEEE Transactions on Control of Network Systems
PWPradeep Chathuranga Weeraddana
Pradeep Chathuranga Weeraddana
In cyberphysical systems, a relevant problem is assigning agents to slots by distributed decisions capable of preserving an agent's privacy. For example, in future intelligent transportation systems, city-level coordinators may optimally assign cars (the agents) to parking slots depending on the cars' distance to final destinations in order to ensure social fairness and without disclosing or even using the car's destination information. Unfortunately, these assignment problems are combinatorial, whereas traditional solvers are exponentially complex, are not scalable, and do not ensure privacy of the agents' intended destinations. Moreover, no emphasis is placed to optimize the agents' social benefit. In this paper, the aggregate social benefit of the agents is considered by an agent-slot assignment optimization problem whose objective function is the fairness among the agents. Due to the problem's complexity, the problem is solved by an approximate approach based on Lagrange duality theory that enables the development of an iterative semidistributed algorithm. It is shown that the proposed algorithm is gracefully scalable compared to centralized methods, and that it preserves privacy in the sense that an eavesdropper will not be able to discover the destination of any agent during the algorithm iterations. Numerical results illustrate the performance and tradeoff of the proposed algorithm compared to the ideal optimal assignment and a greedy method.