Weighted System Throughput Research Articles

CD-MSA: Cooperative and Deadline-Aware Scheduling for Efficient Multi-Tenancy on DNN Accelerators

With DNN turning into the backbone of AI cloud services and propelling the emergence of INFerence-as-a-Service (INFaaS), DNN-specific accelerators have become the indispensable components of cloud inference systems. Due to the conservative “one-task-at-a-time” working mode and deadline blindness of those accelerators, implementing multi-tenancy that aims to improve the cost-effectiveness and meet SLA requirements is intractable. Recent studies including the temporal and spatial approaches, employ manifold scheduling mechanisms and sophisticated architecture innovations to address the challenge. However, these researches either still neglect the deadline awareness or render inevitable and expensive hardware overheads such as switches and storage. In this paper, we present <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Cooperative and Deadline-aware Multi-Systolic-Array scheduling</i> (CD-MSA), a low-cost solution for the cloud inference that utilizes the real time mechanism and task-level parallelism to enable efficient multi-tenancy. Based on our preemptive multi-systolic-array accelerator architecture supporting the simultaneous task co-location, we first construct a fine-grained DNN execution model to lay the groundwork for the lightweight preemption. Second, we design a cooperative, deadline- and laxity-aware scheduler in conjunction with an efficient schedulability test method for better QoS guarantee without introducing additional hardware cost. Finally, to further promote the overall throughput, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dynamic task fusion</i> , a software approach that fuses different tasks into the logically “multi-threading” tasks at runtime. We compare CD-MSA with several state-of-the-art researches across three multi-DNN workloads. The evaluation results show CD-MSA improves the latency-bounded throughput, SLA satisfaction rate and weighted system throughput by up to 62%, 63% and 27%, respectively.

Robust and Secure Resource Allocation for Full-Duplex MISO Multicarrier NOMA Systems

In this paper, we study the resource allocation algorithm design for multiple-input single-output (MISO) multicarrier non-orthogonal multiple access (MC-NOMA) systems, in which a full-duplex base station serves multiple half-duplex uplink and downlink users on the same subcarrier simultaneously. The resource allocation is optimized for maximization of the weighted system throughput while the information leakage is constrained and artificial noise is injected to guarantee secure communication in the presence of multiple potential eavesdroppers. To this end, we formulate a robust non-convex optimization problem taking into account the imperfect channel state information of the eavesdropping channels and the quality-of-service requirements of the legitimate users. Despite the non-convexity of the optimization problem, we solve it optimally by applying monotonic optimization which yields the optimal beamforming, artificial noise design, subcarrier allocation, and power allocation policy. The optimal resource allocation policy serves as a performance benchmark since the corresponding monotonic optimization-based algorithm entails a high computational complexity. Hence, we also develop a low-complexity suboptimal resource allocation algorithm which converges to a locally optimal solution. Our simulation results reveal that the performance of the suboptimal algorithm closely approaches that of the optimal algorithm. Besides, the proposed optimal MISO NOMA system can not only ensure downlink and uplink communication security simultaneously but also provides a significant system secrecy rate improvement compared with the traditional MISO orthogonal multiple access systems and two other baseline schemes.

Resource Allocation and Scheduling in Multi-Cell OFDMA Systems with Decode-and-Forward Relaying

In this paper, we formulate resource allocation and scheduling for multi-cell orthogonal frequency division multiple access (OFDMA) systems with half-duplex decode-and-forward (DF) relaying as a joint optimization problem taking into account multi-cell interference and heterogeneous user data rate requirements. For efficient multi-cell interference mitigation, we incorporate a time slot allocation strategy into the problem formulation. We transform the resulting non-convex and combinatorial optimization problem into a standard convex problem by imposing an interference temperature constraint, which yields a lower bound for the original problem. Subsequently, the transformed optimization problem is solved by dual decomposition and a semi-distributed iterative resource allocation algorithm with closed-form power and subcarrier allocation policies is derived to maximize the average weighted system throughput (bit/s/Hz/base station). Simulation results illustrate that our proposed semi-distributed algorithm achieves practically the same performance as the centralized optimal solution of the original non-convex problem and provides a substantial performance gain compared to single-cell resource allocation and scheduling schemes.

Joint queue control and user scheduling in MIMO broadcast channel under zero‐forcing multiplexing

AbstractThis paper studies the problem of queue control and user scheduling in multi‐antenna broadcast (downlink) systems under zero forcing beamforming transmit strategy. In the system, we assume that the data packet arrives randomly to the buffered transmitter. By taking the broadcast channel as a controlled queueing system, we deduce the property of queue control function that maximizes the weighted system throughput while guarantees the delay fairness among users. We also present a low‐complexity user selection algorithm with the consideration of queue state and channel state together. Simulation results show that the joint queue control and user selection policy can achieve considerable fairness and stability among users. Copyright © 2009 John Wiley &amp; Sons, Ltd.

Utility-based joint power and rate allocation for downlink CDMA with blind multiuser detection

We treat utility-based joint power control and rate allocation in a downlink code division multiple access system, where either the matched-filter or the blind multiuser detection technique is employed at the mobile receivers. Rate allocation is performed on a per-frame basis; whereas within the frame, power control is performed at the symbol basis. A hierarchical rate allocation scheme is proposed, which together with the utility-based power control and the opportunistic fair scheduling scheme, maximizes the instantaneous weighted system throughput and the number of feasible users, and at the same time, guarantees the fairness among all users. In order to apply such a resource allocation framework to systems employing blind multiuser detection, we make use of the recent analytical results on the signal-to-interference-plus-noise-ratio performance of the blind multiuser detectors, assuming multiple spreading codes are used to realize multiple data rates. Our results show that under the utility-based joint power and rate allocation framework, the system capacity is significantly enhanced by using blind multiuser detection and the fairness can be well guaranteed.