Marios Gatzianas

Control of wireless networks with rechargeable batteries [transactions papers]

We consider the problem of cross-layer resource allocation for wireless networks operating with rechargeable batteries under general arrival, channel state and recharge processes. The objective is to maximize total system utility, defined as a function of the long-term rate achieved per link, while satisfying energy and power constraints. A policy with decoupled admission control and power allocation decisions is proposed that achieves asymptotic optimality for sufficiently large battery capacity to maximum transmission power ratio (explicit bounds are provided). We present first a downlink resource allocation scenario; the analysis is then extended to multihop networks. The policy is evaluated via simulations and is seen to perform very well even in the non-asymptotic regime. This policy is particularly suitable for sensor networks, which typically satisfy the asymptotic conditions required by our methodology.

Multiuser Broadcast Erasure Channel With Feedback—Capacity and Algorithms

We consider the N-user broadcast erasure channel with N unicast sessions (one for each user) where receiver feedback is regularly sent to the transmitter in the form of ACK/NACK messages. We first provide a generic outer bound to the capacity of this system; we then propose a virtual-queue-based inter-session mixing coding algorithm, determine its rate region, and show that it achieves capacity under certain conditions on channel statistics, assuming that instantaneous feedback is known to all users. Removing this assumption results in a rate region that asymptotically differs from the outer bound by 1 bit as L → ∞, where L is the number of bits per packet (packet length). For the case of arbitrary channel statistics, we present a modification of the previous algorithm whose rate region is identical to the outer bound for N = 3, when instant feedback is known to all users, and differs from the bound by 1 bit as L → ∞, when the three users know only their own ACK. The proposed algorithms do not require any prior knowledge of channel statistics.

Creating secrets out of erasures

Current security systems often rely on the adversarys computational limitations. Wireless networks offer the opportunity for a different, complementary kind of security, which relies on the adversarys limited network presence (i.e., that the adversary cannot be located at many different points in the network at the same time). We present a system that leverages this opportunity to enable n wireless nodes to create a shared secret S, in a way that an eavesdropper, Eve, obtains very little information on S. Our system consists of two steps: (1) The nodes transmit packets following a special pattern, such that Eve learns very little about a given fraction of the transmitted packets. This is achieved through a combination of beam forming (from many different sources) and wiretap codes. (2) The nodes participate in a protocol that reshuffles the information known to each node, such that the nodes end up sharing a secret that Eve knows very little about. Our protocol is easily implementable in existing wireless devices and scales well with the number of nodes; these properties are achieved through a combination of public feedback, broadcasting, and network coding. We evaluate our system through a 5-node testbed. We demonstrate that a group of wireless nodes can generate thousands of new shared secret bits per second, with their secrecy being independent of the adversarys computational capabilities.

Gain Adaptation Policies for Dual-Hop Nonregenerative Relayed Systems

We examine the performance of a dual-hop nonregenerative system with adjustable relay gain, subject to power constraints. An optimization problem is formulated and solved algorithmically for the binary phase-shift keying bit-error rate utility. The model allows for arbitrary channel statistics. Emphasis is placed on the relation between the optimal solutions obtained when observing the channels of either the first or both hops, as well as the comparison with easily implementable heuristic policies. Numerical results indicate that simple heuristics perform well for a wide range of signal-to-noise ratio (SNR), except for certain high-SNR cases. Finally, the effect of independent channel assumption on system performance is evaluated.

Stable and capacity achieving XOR-based policies for the Broadcast Erasure Channel with feedback

In this paper we describe a network coding scheme for the Broadcast Erasure Channel with multiple unicast stochastic flows, for the case of a single source transmitting packets to N users, where per-slot feedback is fed back to the transmitter in the form of ACK/NACK messages. This scheme performs only binary (XOR) operations and includes special rules for coding packets that ensure instantaneous decodability. Drawing on the results of network stability under statistical overhearing, we provide a stabilizing policy using this coding scheme. Furthermore, we show that, for N = 4 and i.i.d. erasure events, the stability region of such a system effectively coincides with its information-theoretic capacity region, and provide a stabilizing policy that employs this XOR-based scheme.

Feedback-based coding algorithms for broadcast erasure channels with degraded message sets

We consider single-hop broadcast packet erasure channels (BPEC) with degraded message sets and instantaneous feedback regularly available from all receivers, and demonstrate that the main principles of the virtual-queue-based algorithms in [1], which were proposed for multiple unicast sessions, can still be applied to this setting and lead to capacity-achieving algorithms. Specifically, we propose a generic class of algorithms and intuitively describe its rationale and properties that result in its efficiency. We then apply this class of algorithms to three examples of BPEC channels (with different numbers of users and 2 or 3 degraded message sets) and show that the achievable throughput region matches a known capacity outer bound, assuming feedback availability through a separate public channel. If the feedback channel is not public, all users can still decode their messages, albeit at some overhead which results in an achievable throughput that differs from the outer bound by O(N/L), where L is the packet length. These algorithms do not require any prior knowledge of channel statistics for their operation.

Asymptotically Optimal Policies for Wireless Networks with Rechargeable Batteries

We consider the problem of cross-layer resource allocation for single-hop, time-varying, wireless networks operating with rechargeable batteries under general arrival, channel and recharge processes. The objective is to maximize total system utility, as a function of the long-term rate achieved per link, while satisfying energy and power constraints. A policy with decoupled admission control and power allocation decisions is proposed that achieves asymptotic optimality for sufficiently large battery capacity to maximum power ratio (explicit bounds are provided). This policy is particularly suitable for satellite downlinks or sensor networks, which are most likely to satisfy the above condition. An extension to multihop networks is briefly described as well.

Optimal Relay Control in Power-Constrained Dual-Hop Transmissions over Arbitrary Fading Channels

This paper examines the performance of a dualhop, noise-limited, non-regenerative system where the relay gain is allowed to vary as an unknown deterministic function of the channel states, subject to instantaneous (i.e. peak) and average power constraints, the latter directly affecting battery lifetime in wireless systems. A convex optimization problem, in terms of a generic performance metric, is formulated and solved algorithmically for the Signal-to-Noise Ratio (SNR), Shannon rate and BPSK Bit-Error-Rate (BER) utilities. The presented model allows for arbitrary statistics of the fading channels, including correlation between them. Special emphasis is placed on the relation between the optimal solutions obtained when observing the channels of either the first or both hops. The regions in channel state space where zero and full power is allocated are determined in closed form, with numerical results being presented for the Shannon rate and BER utilities. It is observed that, for sufficiently high first hop SNR, monitoring both channels instead of just the first hop can lead to a significant performance increase.

Capacity-achieving encoding for the broadcast erasure channel with multiple users

We consider the N-user memoryless broadcast erasure channel with N unicast sessions (one for each user) where receiver feedback is sent to the transmitter in the form of ACK/NACK messages. We first provide a generic outer bound to the capacity of this system; using concepts from network coding, we then propose a session-mixing coding algorithm applied on specially constructed and maintained virtual queues (at the transmitter side), determine its throughput region and show that it achieves capacity under certain conditions on channel statistics (assuming that instantaneous feedback is known to all users). The algorithm requires no knowledge of channel statistics or future events.

Stable XOR-based policies for the broadcast erasure channel with feedback

In this paper, we describe a network coding scheme for the Broadcast Erasure Channel with multiple unicast stochastic flows, for a single source transmitting packets to N users with per-slot ACK/NACK feedback. This scheme performs only binary (XOR) operations and involves a network of queues, along with special rules for coding and moving packets among the queues, that ensure instantaneous decodability. Additionally, for the scheme to work, one has to specify which packets to select for encoding at each time, based on the received feedback. Contrary to prior work where this packet selection was explicitly specified a priori, we employ a backpressure-type policy that makes the selection based only on queue backlogs. We next provide a stability region outer bound for arbitrary N and erasure patterns and show that this bound effectively coincides with a bound on the systems information-theoretic capacity region (accounting for idle slots). Finally, for N=4 and i.i.d. erasures, we provide a policy that achieves the stability outer bound and employs the proposed XOR scheme using a restricted set of coding rules.