Kimberly M. Wasserman

Dynamic spreading gain control in multiservice CDMA networks

In this paper, we consider a direct-sequence code division multiple access (DS-CDMA) network consisting of a single radio access point and a collection of wireless terminals. The network offers two classes of service: class-1 (real-time) and class-2 (reliable). We are interested in studying the effect of dynamic spreading gain control (SGC) on the dynamics of multiple access interference (MAI), spectral efficiency, and the quality of service (QoS) experienced by each service class. We first consider a time-slotted system in which class-2 terminals operate in a random access fashion. We show that under optimal (throughput maximizing) dynamic SGC: (1) the optimal retransmission probability is equal to one, and (2) the optimal spreading gain increases linearly, or equivalently, the transmission rate decreases inverse linearly, as the MAI level increases. We then model the system as a continuous-time finite-source queueing system with processor sharing, and obtain an explicit (closed-form) expression for the stationary distribution of the number of active class-1 and class-2 terminals, that is, the MAI level. This distribution is used to derive expressions for various QoS measures and define a capacity or admissible region. The results obtained by simulation and analysis are in extremely close agreement. This work contributes to a better understanding of the relationships between QoS, multiple access interference, and allocation of radio resources in DS-CDMA networks.

Distributed power control and spreading gain allocation in CDMA data networks

We study the radio resource allocation problem of distributed joint transmission power control and spreading gain allocation in a DS-CDMA mobile data network. The network consists of K base stations and M wireless data users. The data streams generated by the users are treated as best-effort traffic, in the sense that there are no pre-specified constraints on the quality of the radio channels. We are interested in designing a distributed algorithm that achieves maximal (or near-maximal in some reasonable sense) aggregate throughput, subject to peak power constraints. We provide an algorithm where base stations coordinate in a distributed fashion to control the powers and spreading gains of the users, and show that it converges to a Nash equilibrium point. In general, there may be multiple equilibrium points; however, certain structural properties of the throughput expression can be exploited to significantly trim the search space and induce an ordering on the users in each cell. The numerical results indicate that with these modifications, the algorithm frequently converges in just a few iterations to the throughput maximizing (globally optimal) power and spreading gain allocation.

Optimality of greedy power control and variable spreading gain in multi-class CDMA mobile networks

In this paper, we consider a DS-CDMA mobile network supporting real-time and non-real-time services. We study how the delay tolerance of non-real-time traffic can be exploited to allow both transmission power control and variable spreading gain (transmission rate control) as mechanisms to optimally adapt the received energy per bit to the current channel conditions and efficiently manage the multiple access interference so as to optimize performance. We provide the jointly optimal power and spreading gain allocation strategy that maximizes non-real-time throughput subject to constraints on peak transmission power and maximum interference generated by non-real-time sources. We show that under the optimal strategy, the optimal spreading gains are inverse linear in the signal to interference plus noise ratio (SINR), and transmission power is allocated to the non-realtime sources in decreasing order of channel gain according to a greedy control strategy: The sources with the highest quality channels transmit at maximum power, while the sources with the lowest quality channels do not transmit. The number of non-real-time sources permitted to transmit simultaneously decreases as the peak transmission power increases, and there is at most a 3 dB difference in SlNR between permitting all sources to transmit and permitting only one source to transmit. We also present numerical results comparing the throughput and delay performance of the optimal strategy with other common strategies; the optimal strategy can offer substantial performance gains.

Transmission schemes for time-varying wireless channels with partial state observations

In this paper, we are interested in developing a control algorithm or transmission scheme for wireless data communication over time-varying channels with memory that determines when to attempt a transmission and at what power level, so as to achieve a suitable balance between throughput and energy consumption; and studying the effect of channel memory on the performance, design, and structure of this scheme. The channel state is not directly observable, and thus the transmission decisions must be based on partial or incomplete channel state information, provided to the user by the base station over a feedback channel. More specifically, we assume that the delayed channel state information is provided to the sender at the end of those time slots during which packet transmissions are made. We cast the problem as a partially observable Markov decision process. Despite the difficulty associated with the incomplete channel information, we obtain the optimal transmission scheme, which can be interpreted as a back-off rule: at the end of the transmission, depending on the channel quality during the past time slot, the transmission may be suspended for some time slots, which can be computed in advance. Numerical results are also provided.

Adaptive resource allocation in power constrained CDMA mobile networks

In this paper, we consider a DS-CDMA mobile network supporting real-time and non-real-time services. We study how the delay tolerance of non-real-time traffic can be exploited by allowing both transmission power control and variable spreading gain (transmission rate) control. Two control mechanisms adapt the received energy per bit to the current channel conditions and efficiently manage the multiple access interference so as to optimize performance. We provide the jointly optimal power and spreading gain allocation strategy of non-real-time sources. Our strategy maximizes non-real-time throughput subject to constraints on peak transmission power and protects QoS of real-time-services. We show that under the optimal strategy, the optimal spreading gains are inverse linear in the signal to interference plus noise ratio (SINR), and transmission power is allocated to the non-real-time sources in decreasing order of channel gain according to a greedy control strategy. We also present numerical results comparing the throughput and delay performance of the optimal strategy with other common strategies; the optimal strategy can offer substantial performance gains.

Fugue: time scales of adaptation in mobile video

Providing interactive video on hand-held, mobile devices is extremely difficult. These devices are subject to processor, memory, and power constraints, and communicate over wireless links of rapidly varying quality. Furthermore, the size of encoded video is difficult to predict, complicating the encoding task. We present Fugue, a system that copes with these challenges through a division along time scales of adaptation. Fugue is structured as three sperate controllers: transmission, video and preference. This decomposition provides adaptation along different time scales: per-packet, per-frame, and per-video. The controllers are provided at modest time and space costs compared to the cost of video encoding. We present simulations confirming the efficacy of our transmission controller, and compare our video controller to several alternatives. We find that, in situations amenable to adaptive compression, our scheme provides video quality equal to or better than the alternatives at a comparable or substantially lower computational cost. We also find that distortion, the metric commonly used to compare mobile video, under-values the contribution smooth motion makes to perceived video quality.

Energy efficient data communication over fading channels

We develop a control algorithm or transmission scheme for wireless data communication over time-varying channels with memory which determines when to attempt a transmission so as to achieve a suitable balance between throughput and energy consumption; and we study the effect of channel memory on the performance, design, and structure of this scheme. The channel state is not directly observable, and thus the transmission decisions must be based on ACK/NAK information provided over a feedback channel. We cast the problem as a partially observable Markov decision process, and derive the optimal transmission scheme. Numerical results are also provided comparing performance of the optimal scheme with some other existing mechanisms.

Estimation of network link loss rates via chaining in multicast trees

Of increasing importance is estimation of internal link parameters in communications networks. Multicast probes are a way to gather statistics about internal links from edge node measurements. The problem of estimating link loss probabilities for a multicast distribution tree is examined. Our model assumes loss statistics are distributed to session participants by a network protocol such as RTCP. We propose a decentralized algorithm for ML estimation of the link loss probabilities in a chain of nodes rooted at the source node of the multicast distribution tree and terminating at a given leaf. An expression for the Cramer-Rao bound and an approximate form for the probability distribution function of the estimator are given. The performance of the algorithm is evaluated using computer simulations for a bottleneck detection application.

On Mutually Interfering Parallel Servers Subject to External Disturbances

This paper considers a continuous-time non-Markovian parallel queueing system subject to external disturbances. The servers are mutually interfering in that their service rates are nonlinearly interdependent functions of the controls applied by the servers, and external discrete-valued continuous-time random disturbances. At certain time epochs, namely, every ? time units, the servers may adjust their service rates by changing the values of their controls; however, the system may change its state several times between successive decision epochs. The stability region of the system is established and a service rate control policy p* is provided, where an arrival rate vector in the interior of the region is sufficient for stability under p*, and a vector in the closure is necessary for stability under any policy. The stability region depends on ? and the variations of the disturbances between decision epochs, and p* does not require knowledge of the arrival rates. The stability region is not in general monotonic in ?, but under perfect continuous control (? = 0) the stability region is a superset of that under ? > 0. This queueing model captures essential features of resource allocation and stochastic control problems encountered in a number of telecommunication, transportation, and manufacturing systems.

Effect of channel memory on retransmission protocols for low energy wireless data communications

The decision to attempt or suspend transmission is cast as a stochastic control problem with imperfect state information. By employing a dynamic programming formulation, the tradeoff between high throughput and energy efficiency is resolved in a flexible cost structure over which the optimization is performed. The optimal policy is derived and shown to be a threshold rule that varies with the memory present in the error process. A suboptimal implementation of the policy indicates that the protocol always favors attempting transmissions when the memory is low. As the channel memory increases, the protocol suspends transmission for longer durations after a packet failure.