Packet Scheduling Algorithms for Mixed Traffic Flow over HSDPA

Only best-effort traffic was supported by traditional packet switching networks, but newer applications need network services that allow an end-client to transport data with performance guarantees. The choice of packet-scheduling algorithm used at switching nodes is crucial in providing Quality of Service (QoS). This paper proposes a downlink packet scheduling algorithm for QoS provisioning in High Speed Downlink Packet Access (HSDPA) based UMTS networks. In HSDPA, a single logical channel is shared amongst multiple contending users. Besides better link utilization, scheduling disciplines seek to achieve fair allocation of this shared resource. However, these two criteria can potentially be in conflict. The scheduling algorithm proposed in this paper, Wireless Fair High Speed Scheduling, addresses this conflict. The basic idea is to allocate the wireless channel fairly amongst users, even during short time scales, according to their data rate requirements. Simulation results show that the proposed scheduling algorithm provides better link utilization, as well as long-term and short-term fairness amongst users, even under heavy load conditions.

This paper proposes a novel packet scheduling algorithm for delay constraint traffic on high speed down link packet access (HSDPA) and also a packet scheduler for this algorithm. In this packet scheduling algorithm, we consider both the delay constraint and the throughput for each traffic class while maximizing the system capacity. Considering the delay constraint and the throughput simultaneously, our method can guarantee different quality of service (QoS) for each traffic class. Finally, in order to demonstrate the performance of our proposed algorithm, we compare our algorithm with the proportional fairness (PF) method and round robin (RR) method by system simulations. Numerical results show that system performance can be significantly improved employing our algorithm compared to that with the PF or RR method.

HSDPA (high-speed downlink packet access), introduced in WCDMA release 5, provides a high-bandwidth shared channel with short transmission time interval (TTI). The short TTI together with appropriate scheduling enable HSDPA to support efficient multiplexing of traffic. We explain the performance of four scheduling algorithms when transmitting a traffic mix consisting of both conversational (VoIP) traffic and background (Web) traffic over the high-speed downlink shared channel (HS-DSCH) of HSDPA. We consider both cell throughput and user satisfaction. The proportional fair (PF), the maximum rate (MR) scheduler and two extended versions of MR, are tested for different VoIP scheduling delay budgets and varying load. To understand the behaviour of the schedulers, we use the ns-2 simulator extended with a model of HS-DSCH to simulate a mixed traffic scenario. Our results show that a scheduler that gradually increases the VoIP priority and considers the user's current possible rate, performs well. A more drastic increase in VoIP priority is however needed when the delay budget is short. Furthermore, attempting to uphold quality for both VoIP and Web traffic makes the system sensitive to overload situations.

LTE-Advanced offers significantly higher data rates than the legacy system. The carrier aggregation (CA) technology allows scalable expansion of effective bandwidth provided to user equipment (UE) through simultaneous utilization of radio resources across multi-component carriers. In this paper we propose a new packet scheduling (PS) criterion algorithm that satisfies the fairness among the different kinds of UEs by designing a weighting factor to proportional fair (PF) packet scheduling (PS) algorithms, while enhancing their throughput performance. The proposed PS algorithm is implemented and validated in a PS module for LTE/LTE-Advanced via system level simulations. Results show that PS-enhanced algorithm achieve higher throughput for both LTE and LTE-Advanced UEs.

High speed downlink packet access (HSDPA) is a packet-based data service that supports downlink data rates of up to 14 Mbps in the WCDMA (wideband code division multiple access) system. In HSDPA, the packet scheduler is a key element for high-speed and efficient transmissions. In addition, hybrid automatic request (HARQ) plays a significant role in improving system performance. We propose an enhanced scheduling algorithm combined with HARQ to reduce transfer delay without the degradation of system throughput.

In order to improve support for higher speed data, 3GPP has developed an evolution of UNTS based on WCDMA known as HSDPA(high speed downlink packet access). The target of HSDPA is to provide service for not only voice user but also higher speed data users. For HSDPA, there are two important parameters, which are the rate ratio and the power ratio between higher speed data signal and voice data signal. In this article, we deduced two useful conclusions, which are based on the research of the relationship between the rate ratio and the power ratio. One of the two conclusions is that the transmission power is in positive proportional to data rate when the user numbers of higher speed data is high, which would result in dramatic interference of multi-user. The other is that the same capacity can be achieved by only using half transmission power when there is one higher speed data user over one transmission time interval. Based on the two conclusions, the feasibility of HSDPA is studied. We found that it is impossible for HSDPA to provide simultaneously service for multiple high-speed data users and voice users. A feasible CDMA/TDMA reform scheme is proposed, which can achieve the design goal of WCDMA HSDPA. The reform scheme employs the method of time domain multiplexing(TDM) to realize the transmission of high speed data over a special CDMA channel. The reform scheme can increase dramatically downlink capacity, improve spectral efficiency, and also reduce system cost by employing simple interference eliminator.

The 3G UMTS is currently undergoing inherent changes. The expected release 5&6 of UMTS will contain a new set of features known collectively as high speed downlink packet access (HSDPA). In this paper, we propose two packet scheduling algorithms for HSDPA, shortest queue first (SQF) and longest queue first (LQF), to provide priority scheduling between services of different quality of service (QoS) classes and fairness between users within the same class. Simulation results show that the proposed algorithms meet QoS requirements of different traffic classes, with the SQF algorithm having a better performance compared to the LQF algorithm.

Simple wireless scheduling techniques, such as maximum carrier-to-interference ratio (Max. C/I) or round robin (RR), fail to take into account all aspects of quality of service (QoS) provisioning. The fundamentals of evolutionary computation (genetic algorithms) and conventional wireless scheduling techniques are combined and the weaknesses of the existing known techniques are exploited to propose a novel hybrid genetic packet scheduler (HGPS) for high speed downlink packet access (HSDPA). The proposed HGPS outperforms the Max. C/I packet scheduler in terms of total delivered throughput within low delay thresholds.

High speed downlink packet access (HSDPA) was originally introduced in release 5 of WCDMA to boost data rates of packet switched traffic in the downlink. With increasing demand for streaming services and recent efforts to provide mobile TV, the standard has also been considered for delivering streaming services over cellular networks. While much attention has been paid to optimizing transmission of non real time application, this paper aims at providing a comprehensive study of particular requirements and performance aspects of transmission of streaming services over HSDPA. We compare performances of different scheduling algorithms in pure streaming scenarios for constant bit rate as well as variable bit rate streams and different streaming rates. We further study the impact of code multiplexing of different streaming users on the system performance and examine the system behavior when both streaming and best effort services are present.

The provision of data services in 3G systems requires spectral efficiency solutions to increase the capacity of the radio access network as well as support for high data rates. In this context, the 3GPP introduces a new feature in the release 5 specifications denominated high speed downlink packet access (HSDPA). The HSDPA concept appears as an umbrella of features to improve both user and system performance. The concept includes a reduction of the transmission time interval (TTI) to 2 ms, an adaptive modulation and coding (AMC) scheme, and a fast physical layer hybrid ARQ mechanism. In this paper the objective is to analyze the HSDPA performance for streaming, interactive, and background services. Relevant characteristics such as the QoS demands of the different traffic classes, the throughput and delay characteristics has been analyzed and compared with shared channel. We also propose a new packet-scheduling scheme for HSDPA transmission, which improves the overall performance and fairness among the packet users.

In order to provide the growing demand of data services, several wireless standards are evolving to support packet data more efficiently. In particular 3rd generation partnership project (3GPP) is studying a new access technique called high speed downlink packet access (HSDPA) to support UMTS users with high speed data services. The present paper is focused on scheduling problems for video streaming services in the HSDPA. Two classical scheduling algorithms (weighted first queuing and earliest deadline first) are applied to this access technique and a new algorithm, based on an hybrid procedure between the above, is also designed. Furthermore two strategies for radio resource allocation are introduced: frame-by-frame assignment and fast scheduling, respectively. A software simulation using OPNET modeler is performed in order to compare and find out benefits and drawbacks in the application of each above mentioned strategy.

This paper presents an analytical model of the cell and user throughput of high-speed downlink packet access (HSDPA) that can be used to compare the performance and efficiency of various schedulers. The focus is on the proportional fair (PF), score-based (SB), and Max C/I schedulers. The radio channel is assumed to be a composite multipath/shadowing channel, with the path amplitude following a Nakagami distribution. The effect of the correlation between paths on the HSDPA throughput performance is taken into account in the modeling. The analytical model is confronted with a simulation using the NS2 network simulator and shown to provide reasonably close results to the simulation in view of the modeling assumptions.

Fast packet scheduling is a key feature of packet-based data services which are 3G/4G wireless networks such as Enhanced Data rates for GSM Evolution, High-Speed Downlink Packet Access (HSDPA), and High Data Rate. In this paper, we propose Anticipated Proportionally Optimal (APO), a new packet scheduling scheme with Quality of Service (QoS) support for hybrid of real and non-real-time services over HSDPA networks. The APO scheduling algorithm is based on the effective channel anticipation model. In contrast to the traditional schemes, the proposed method is implemented based on a cyclic non-work-conserving discipline. Simulation results indicate capability of proposed scheme to maximize the channel usage efficiency compared to other exist scheduling methods. Results also present fair tradeoff between channel capacity and desired QoS for real-time services.

In Wideband Code Division Multiple Access (WCDMA) networks the Signaling Radio Bearer (SRB) is transmitted downlink to inform the user i.a. when it needs to make a handover. When transmitting e.g. Voice over IP (VoIP) on High Speed Downlink Packet Access (HSDPA) SRB is mapped on HSDPA user data channel, high-speed downlink shared channel (HS-DSCH) in order to maximize voice capacity. The lack of support for soft handover (SHO) of HS-DSCH may compromise the reception of SRB and thus also a possible incoming serving HSDPA cell change command especially at the cell borders. This can lead to serious VoIP service degradation due to prolonged camping on a poor cell and in the worst case even to call dropping, if the cell change command is not received before VoIP call drop criterion is met. This paper considers the performance of SRB transmission on source cell HS-DSCH during a handover process. Also a strategy of transmitting the serving cell change command from the target cell is analyzed. The study is done by the means of fully dynamic system simulations and they consider both macro-cell and manhattan scenarios. The results show that a relatively low SRB/handover error rate is experienced in a macro cell scenario but in a more challenging manhattan scenario the error rate is higher. Transmitting the serving cell change command from the target cell HS-SCCH results in similar handover failure rates in manhattan scenario than using SRB on HS-DSCH in macro-cell scenario when static power allocation for HS-SCCH is in use.

Component carriers (CCs) in a Long Term Evolution - Advanced (LTE-A) system are limited by maximum 100 radio bearers (RBs) on a 20 MHz CC due to a fixed 15 KHz subcarrier bandwidth. However it is able to be changed in the future, so that a study regarding heterogeneous CCs which have diverse RBs above 100 on a CC is needed at this moment. To support non-contiguous carrier aggregation (CA) with the these conditions, a heterogeneous joint CC packet scheduling algorithm interacting with a radio resource control (RRC) is considered in this paper. The performance of the joint CC packet scheduling algorithm on heterogeneous CCs with different channel characteristics has been analyzed and compared through computer simulations as well. According to the results, we found that an 800 MHz CC exhibits approximately 4 dB better channel gains than a 2.1 GHz CC. We also obtained a capacity increase of about 15-40% for a joint CC maximum rate (MR) algorithm and about 2-28 % for a joint CC proportional fair (PF) algorithm when the number of users is increased. On the other hand, in a round-robin (RR) scenario, we were unable to notice any differences between the independent CC and joint CC packet scheduling policies. Furthermore, looking at the impacts of different amounts of RBs on 800 MHz + 2.1 GHz CCs, we can conclude that a one-to-one RB ratio with respect to the MR, PF, and RR algorithms has a better performance compared to other cases, and thus, owing to the better channel gain, the smaller the RBs of the 800 MHz CC, the better the performance achieved.