Keyvan Zarifi

Decentralized Relay Selection Schemes in Uniformly Distributed Wireless Sensor Networks

We study three relay selection schemes for uniformly distributed wireless sensor networks: 1) optimal selection where the relays that maximize the signal-to-noise ratio (SNR) at the destination are selected, 2) geometry-based, which is based on selecting the closest nodes to the source, and 3) random selection in which the nodes are selected randomly from a certain neighborhood of the source. In all schemes, we assume that all relays operate in the amplify-and-forward mode and transmit with equal average powers and each relay has only access to its backward channel and location. For each relay selection strategy, we propose a decentralized protocol whereby proper nodes choose to act as relays without requiring any central coordinating entity or any inter-node information transfer. We derive expressions for the average SNR at the relays and destination while assuming that the source-node distances and the inter-terminal channel links are completely random. We show that, for all proposed schemes, the SNR variance at the destination converges to zero as the number of relays increases. While each selection scheme has its pros and cons, we derive a sufficient condition under which the average SNR at the destination becomes independent of the selection scheme employed.

Distributed beamforming for wireless sensor networks with random node location

Due to the unsupervised nature of wireless sensor networks (WSNs), intensive communications are required among the selected nodes to reach a consensus and synchronize prior to entering a distributed beamforming (DBF) procedure. Therefore, a sensible approach to select the nodes should not only take into account the required beampattern, but also should aim to preserve the inter-node connectivity and the network energy. We show for a uniformly distributed WSN that when the nodes are selected from a ring of proper radii, the resulting beampattern mainlobe is narrower compared to that of the classical DBF technique proposed in [1]. At the same time, our proposed technique may preserve a substantial amount of network energy and reduce the probability of network disconnectivity. Directivity of the proposed DBF technique is analyzed and an extension of the technique to a multi-ring case is presented. It is shown that the sidelobe peaks can be considerably decreased if the nodes are selected from multiple concentric rings.

Radio access virtualization: Cell follows user

Virtual radio access (VRA) technology wherein groups of cooperative transmit points (TPs) form virtual TPs (VTPs) to serve user equipments (UEs) continue to be a thriving subject of research in future generations of wireless networks. In this paper, we propose a technique that uses UE-centric metrics to provide multiple partitions of a wireless network into VTP sets. Our technique guarantees that all UEs enjoy a required gain in at least one VTP; effectively eliminating the edge UE experience in the network. To further enhance the performance of the proposed VRA technique in practical scenarios wherein there is a large load imbalance in the network, we also introduce a new concept of soft UE-TP association in which each UE is partially associated with multiple TPs. The use of our soft association concept when forming VTP sets facilitates load-balancing among various TPs. Finally, a technique is also offered to select the best VTP set at each scheduling resource unit. Numerical simulations are used to demonstrate the performance of our virtualization techniques.

An approach to distributed implementation of cooperative amplify-and-forward beamforming in wireless sensor networks

We consider a SINR-optimal dual-hop amplify-and-forward beamformer and modify it such that the resulting beam-former well-approximates its original counterpart and, further, lends itself to a distributed implementation in wireless sensor networks wherein each node is unaware of the locations and the forward channels of all other nodes. We then derive the average beampattern expression of the proposed beamformer and analyze its properties.

Distributed processing techniques for beamforming in wireless sensor networks

A main task in distributed beamforming (DBF) techniques for wireless sensor networks (WSNs) is to maximize the received signal power at the access point (AP) while inflicting small interfering effect on unintended receivers. When the DBF nodes are unaware of the directions of unintended receivers, interference at the latter receivers may be substantially reduced by forming a beampattern with a narrow mainlobe that is pointed towards the AP. However, such an approach requires the DBF nodes to be sporadically scattered over a large area and, hence, increases the probability of the loss of their inter-connection. Assuming that nodes are uniformly distributed in the network, we show how the DBF nodes can be intelligently selected to ameliorate the network disconnectivity problem. In turn, when the directions of unintended receivers are known, one may aim to apply the so-called null-steering beamforming approach to effectively nullify the received power at those directions. However, it can be shown that implementing a null-steering beamformer in WSNs requires each node to be aware of the locations of all other nodes in the network; a requirement that opposes the distributed nature of WSNs. For such a scenario, we approximate the null-steering beamformer with another beamformer that is amenable to a distributed implementation.

Progressive hybrid greyfield wireless access virtualization: Graph-optimized dynamic utility tradeoffs between cloud, fog, and legacy RANs

In this paper, we develop a new dynamic utility for wireless access virtualization (WAV) optimization embodying highly-dimensional time-varying multi-criteria metrics (i.e., CAPEX and OPEX costs, QoS or QoE, multi-tier and/or multi-RAT HetNets, etc.) that gauge the best deployment and viability scenarios of cloud (C)- and fog (F)-RANs within legacy networks. Exploiting the powerful tool of graph theory, we devise a progressive greyfield WAV strategy that optimizes our dynamic utility through an efficient combination of C- and F-RANs. This strategy is able to readjust very quickly to any changes in existing or new constraints as they evolve or occur in time, respectively. The resulting optimized hybrid RAN deployment outperforms both the greenfield and the pre-planed greyfield “turnkey” WAV strategies.

Two-tier distributed and open loop multi-point cooperation using SCMA

The fifth generation of cellular wireless networks known as 5G is based on user-centric non-cellular concept where the users are surrounded by many network nodes cooperating to serve the users providing a “cell-center” experience throughout the network. Traditional multi point cooperation techniques often rely on centralized coordination and require different and often stringent requirement on the central controller, backhaul capacity and overall network synchronization. A novel two tiered, open loop and distributed cooperation technique is proposed in this paper where the lower tier with fixed or slowly changing parameters and preferably using SCMA provide a ubiquitous performance to mobile users and users exposed to multiple network nodes, while the higher tier provides service to the users close to the network nodes and maintain the overall network throughput. Users scheduled by the lower tier signaling use joint detection techniques from multiple network nodes. Other users scheduled to the higher tier jointly decode the lower tier signal from one or multiple network nodes before proceeding with the detection of their intended signal. The proposed algorithm is based on distributed scheduling and imposes limited requirements on the central controller and backhaul capacity and does not require stringent network time and frequency synchronization among network nodes. Unlike traditional cooperation techniques, the proposed method is open loop and requires very limited feedback and signaling overhead. Performance evaluations show that the proposed technique provides significant network coverage enhancement especially for high speed mobile users.

Dummy users; Network-aware cooperative terminals in wireless access networks

Cooperation among network nodes at one side and terminal nodes on the other side are the two main enablers of radio access virtualization to establish scalable and dynamically adjustable data pipes to the users in the 5G mobile networks. User cooperation requires careful design to address cooperation incentive, battery life, user discovery, security and privacy. Dummy user, a network-aware terminal, can facilitate the implementation of user cooperation and leverage its benefits. Dummy users can offer a low cost alternative for the enhancement of the network performance to address persistent or temporary demand such as a hotspot created by an outdoor festival. Proper deployment of dummy users provides fairness and ubiquity of the network while enabling many advanced transmission and reception schemes that require more accurate channel state knowledge at the network.

Asymptotic Analysis and Design of Multiuser Cooperative DS-CDMA Systems

The performance of a cooperative multiuser direct-sequence code-division multiple-access (DS-CDMA) system is analyzed in the asymptotic regime where both the spreading codes and the number of users grow large with the same ratio. A simple signal-to-interference-plus-noise ratio (SINR) expression is derived that is independent from the spreading codes and explicitly accounts for the effects of the multiple-access interference (MAI) and the relay noise. The so-obtained SINR expression is then computed based entirely on the available local information. The results obtained above are then used to optimally design the cooperative system. In particular, it is shown how the amount of cooperation between each collaborating pair can be adjusted to simultaneously achieve a preassigned target SINR for both users. Based on the local information, the globally optimal amount of the relay power is also obtained that maximizes the achieved SINR at the access point.