Patrick Hosein

QoS control for WCDMA high speed packet data

Wideband CDMA Release 5 is expected to support peak downlink bit rates of 10 Mbps for use with bandwidth intensive data and multimedia applications. Such rates are achieved through fast link adaptation, fast Hybrid ARQ and fast scheduling over a shared forward link packet data channel. Total throughput can be maximized by having the frame scheduler take into account the instantaneous radio conditions of users and serving users during their good radio condition periods. This results in high user diversity gains. However in order to support QoS guarantees, it will sometimes be necessary to serve users experiencing bad radio conditions in order to maintain their requested QoS levels. We present a flexible algorithm that provides user QoS guarantees while at the same time achieving some user diversity gains.

Capacity of packetized voice services over time-shared wireless packet data channels

In traditional CDMA wireless networks, real-time services, such as circuit-switched voice, are transported over the air via synchronous channels because of their stringent delay requirements. In the third generation networks (3G), IxEV-DV [A.Soong et al., 2003] and HSDPA [E. Dahlman et al., 1998], an additional time shared channel was introduced on the forward link to support data services for which delay requirements are less stringent. The IxEV-DO [P. Bender et al., 2000] standard provides a single time-shared forward link channel and no synchronous channels since it was designed primarily for data services. It has recently been suggested that certain voice services (e.g.. Voice over IP (VoIP) and push-to-talk (PTT)), can be efficiently transported over such time-shared channels since they have less stringent delay requirements compared to circuit-switched voice. In this paper we investigate the capacity of VoIP users over these time-shared channels and investigate the sensitivity to various base station (BS) and mobile station (MS) design parameters, algorithms and features. Note that detailed simulations of each standard is not provided but rather a comparative approach is used whereby we investigate specific features of each standard. Our focus is on the forward channel since code division multiplexing is performed in the reverse link and hence comparable user capacities are achieved.

Scheduling of VoIP traffic over a time-shared wireless packet data channel

In traditional wireless networks, real-time services (e.g., voice) are transported over synchronous channels because of their delay sensitivity while data is transported over asynchronous channels because of its burstiness. It has recently been proposed that even real-time services can be efficiently transported over asynchronous channels, such as the time-shared channels supported on the forward link of 3G networks such as 1xEV-DO (Bender, P. et al., 2000), 1xEV-DV (Soong, A. et al., 2003) and HSDPA (Dahlman, E. et al., 1998). In order to provide acceptable user performance for services such as voice over IP (VoIP), while at the same time providing high user capacity, the allocation of resources to users must be carefully managed. In particular, the scheduler that allocates time slots to users must be carefully designed. We formulate the problem of scheduling VoIP users over a generic time-shared wireless packet data channel. We then propose a scheduling algorithm that achieves our stated objectives.

Optimal frequency of Walsh mask broadcast for forward high-speed wireless packet data channels

The 1/spl times/EV-DV 3G wireless network system is able to simultaneously support traditional code division multiplexed voice and circuit-switched data channels in addition to one or two time shared high speed packet data channels. In practice, voice calls will be provided with higher priority than data calls. Therefore, radio resources (e.g., power and Walsh codes) will be first given to voice connections and any unused resources will be dynamically assigned to the high speed packet data channel. Because of Walsh code fragmentation, it may be possible that Walsh codes not being used for voice and circuit-switched data connections cannot be used by the packet data channel. In order to make use of such codes, the Walsh code mask must be broadcast in the sector. Such a broadcast results in increased sector throughput due to the additional Walsh codes available to the forward packet data channel (F-PDCH) but it also requires forward link resources which reduces throughput. We determine the broadcast frequency that optimizes this tradeoff.

Dynamic power headroom threshold for mobile rate determination in a CDMA network

Until recently, the focus in 3G networks has been on the forward link since this typically requires a higher bandwidth than the reverse direction. We focus on the reverse link since this also needs to have a sufficiently high bandwidth to support reverse link bandwidth intensive applications. H-ARQ can be used to improve throughput efficiency since errors are corrected at the physical instead of at the MAC layer. However, when this is used, the mobile must take into account possible future re-transmissions when deciding on the transmission rate of the first transmission. This requires reserving enough transmission power headroom in case additional power is needed for a subsequent re-transmission. This depends on the variability of the reverse channel conditions which in turn depends on the mobility of the mobile as well as the reverse link load of the sector. In this paper, we propose setting the power headroom dynamically instead of using a static value as is presently done. Simulation results demonstrate the effectiveness of our approach.

A CLASS-BASED ADMISSION CONTROL ALGORITHM FOR SHARED WIRELESS CHANNELS SUPPORTING QOS SERVICES

Third Generation (3G) wireless networks all support a shared high-speed downlink channel for efficient transport of data. These channels use several common techniques (fast link rate adaptation, Hybrid ARQ, fast cell selection and fast scheduling) to improve performance. In addition, future 3G networks must provide the necessary framework to allow operators to offer Quality of Service (QoS) based applications. Service differentiation will also become necessary so that operators can provide differing service levels based on user classes. Pricing will be based on these user classes so that the operators can optimize their revenue. Typically, the shared forward link channel will be the bottleneck resource and hence usage of this resource must be optimally managed. This means that admission controls must be used to ensure that admitted users receive their QoS guarantees for the duration of their connection. If Class-Based Grade of Service (GoS) is offered then the admission control algorithm must also provide class dependent connection blocking rates. In this paper we present a framework that allows an operator to provide Class-Based connection admissions.

On the tradeoff between throughput and fairness on the reverse link of a 3G CDMA network

Third generation (3G) networks will provide support for high speed applications to its users in both the forward and reverse directions. In the forward direction, this is achieved by time sharing users over one or more high speed packet data channels. Users are scheduled based on their application needs as well as the mobile operators objectives. On the reverse link, two options are possible for serving users, rate control and scheduling. Scheduling provides fine control of reverse link resources but at the expense of complexity and signaling overhead. Two forms of rate control are possible, common rate control and dedicated rate control. Dedicated rate control is similar in concept to scheduling, but is limited to one rate level change per user. In common rate control, each mobile independently determines its reverse link rate using load information that is broadcast by the base station on a periodic basis. The algorithm used by each mobile to determine its transmission rate should again be determined based on the needs of the application as well as the operators overall objectives. One option is to be completely fair and provide all mobiles (as much as is physically possible) with identical transmission rates (i.e., the performance objective used for voice users). Another option is to allow mobiles to use rates based on their reverse link channel conditions (the proportionally fair performance objective used in the forward packet data channel). We compare the throughput and fairness of these two approaches.

Optimality conditions for throughput maximization on the reverse link of a CDMA network

We consider the information-theoretic sum capacity of the reverse link of a CDMA wireless network. In particular we consider the problem of maximizing the throughput of this channel. The decision variables are the transmission powers that are allocated to the mobiles in the network. The major contribution of this paper is the derivation of thresholds that determine when it is: (a) optimal to serve only a single user; and (b) optimal to simultaneously serve multiple users. We also derive the optimal power allocations for the case where users are provided guaranteed minimum rates. Finally we provide a numerical example where we compare the gains of scheduling subsets of users at a time instead of all users simultaneously while maintaining some degree of fairness.

Simple fairness metric for time-shared wireless data networks

Third generation networks (e.g, 1xEV-DO, 1xEV-DV and WCDMA) all improve the efficiency of data transmissions by taking advantage of the fact that data is less sensitive to delays than voice. Therefore, instead of serving a data stream synchronously, they allow for bursty servicing of each data stream. In the forward link, each data stream burst is typically scheduled when the corresponding mobile is in favorable radio conditions to maximize radio resource efficiency. This results in a trade-off between total sector throughput and fairness among users within the sector. Total sector throughput is increased by serving users in good average radio conditions while fairness is achieved by providing all users with identical throughputs. Naturally, these two objectives conflict and some trade-off must be made between them. While sector throughput is easily measured, the fairness of different schedulers have typically been compared by visual inspection of the cumulative distribution function of user throughputs. We recommend using a more objective criterion for comparing schedulers, namely the Jain fairness index. We illustrate how this metric accurately reflects fairness and provide examples showing the throughput/fairness trade-off.

Integrated scheduling and buffer management for 3G wireless forward packet data channels

Third generation (3G) wireless networks have been designed to provide high speed data services on both forward and reverse links. In addition to increased data capacity they also support QoS (quality of service) guarantees for user applications. Users may run multiple applications (service instances) simultaneously and QoS guarantees must be assured for one or more of them. In order to maintain these QoS guarantees for each users service instance, the radio resources assigned to the mobile must be adjusted in real time. However, the forward packet data channel in 3G systems are time-shared among all mobiles. Furthermore, the forward link channel conditions of the mobile changes over time. This implies that the forward link throughput available to the mobile is stochastic. In this paper, we address the linkage between the application demands (i.e., based on the QoS attribute guarantees) and the available channel capacity (i.e., resources assigned by the scheduler) and present a simple solution to the problem.