ABSTRACT Heterogeneous architectures are vastly used in various high performance computing systems from IoT‐based embedded architectures to edge and cloud systems. Although heterogeneous architectures with cooperation of CPUs and GPUs and unified address space are increasingly used, there are still a lot of open questions and challenges regarding the design of these architectures. For evaluation, validation and exploration of next generation of heterogeneous CPU–GPU architectures, it is essential to use unified heterogeneous simulators for analyzing the execution of CPU–GPU workloads. This article presents a systematic review on challenges of heterogeneous CPU–GPU architectures with covering a diverse set of literatures on each challenge. The main considered challenges are shared resource management, network interconnections, task scheduling, energy consumption, and programming model. In addition, in this article, the state‐of‐the‐art of heterogeneous CPU–GPU simulation platforms is reviewed. The structure and characteristics of five cycle‐accurate heterogeneous CPU–GPU simulators are described and compared. We perform comprehensive discussions on the methodologies and challenges of designing high performance heterogeneous architectures. Moreover, for developing efficient heterogeneous CPU–GPU simulators, some recommendations are presented.

Summary High performance system‐on‐chip (SoCs) designs have led to high‐density integrated circuits using field programmable gate arrays (FPGAs) for rapid prototyping and reconfigurable digital circuits. Using FPGA reconfigurability, it is possible to design a configurable network‐on‐chip (NoC) for different applications. NoC architectures provide efficient communication infrastructures for implementing very large SoCs. In this article, we propose HiFMP, a high‐performance FPGA‐based multicrossbar prioritized NoC router. The aim followed by the proposed router is designing a low‐power NoC router with high performance in terms of energy‐efficiency, network throughput, area, and latency for efficient FPGA realization. HiFMP is a parameterizable router, and is effectively used for an FPGA‐based NoC with mesh topology. Performance evaluations include network‐level analysis and hardware exploration; the results demonstrate the effectiveness and high performance of HiFMP in terms of latency, throughput, power consumption, and area, comparing with the existing related architectures.

We present Duo-MAC, an asynchronous cascading wake-up scheduled MAC protocol for heterogeneous traffic forwarding in low-power wireless networks. Duo-MAC deals with energy-delay minimization problem and copes with transmission latency encountered by Today's duty-cycled protocols when forwarding heterogeneous traffic types. It switches, according to the energy and delay requirements, between Low Duty cycle (LDC) and High Duty Cycle (HDC) operating modes, and it quietly adjusts the wake-up schedule of a node according to (i) its parent's wake-up time and (ii) its estimated load, using an effective real-time signal processing linear traffic estimator. As a second contribution, Duo-MAC, proposes a service differentiation through an improved contention window adaptation algorithm to meet delay requirements of heterogeneous traffic classes. Duo-MAC's efficiency stems from balancing between the two traffic award operation modes. Implementation and experimentation of Duo-MAC on a MicaZ mote platform reveals that the protocol outperforms other state-of-the-art MAC protocols from the energy-delay minimization perspective.

Clustering techniques are proposed for developing performance of wireless sensor networks from several points of view. Multimedia nodes have different sensing region than scalar sensors that usually sense a circle or polygon regions centred by the node, and thus traditional clustering methods do not satisfy a Wireless Multimedia Sensor Network (WMSN). In this paper, a novel formation algorithm, Multi-Cluster-Membership (MCM), for multimedia node clustering is proposed. The proposed algorithm establishes clusters overlapping each other and the common members of clusters are coordinated to make cooperation among clusters. Therefore, MCM offers not only intra-cluster cooperation but also inter-cluster cooperation through coordinating members and also clusters. This differs from other coordinating ways in the fact that nodes may participate sensing in multiple clusters. We show the enhancement in power conservation efficiency of network resulted from applying MCM with respect the Single-Cluster-Membership (SCM) algorithm.

Wireless Multimedia Sensor nodes sense areas that are uncorrelated to the areas covered by radio neighbor sensors. Thus, node clustering for coordinating multimedia sensing and processing cannot be based on classical sensor clustering algorithms. This paper presents a clustering mechanism for Wireless Multimedia Sensor Networks based on overlapped Field of View (FoV) areas. Today, for random deployments, dense networks of low cost, low resolution and low power multimedia nodes are preferred than sparse cases of high cost, high resolution and high power nodes. Overlapping FoVs in dense networks causes wasting power of system because of redundant sensing of area. The main aim of the proposed clustering method is energy conservation and prolonging network lifetime. This aim is achieved through coordination of nodes belonging to the same cluster in assigned tasks, avoiding redundant sensing or processing.

A critical aspect of applications with wireless sensor networks is network lifetime. Sensing and communications consume energy particularly in wireless multimedia sensor networks (WMSN) due to huge amount of data generated by the multimedia sensors. Therefore, judicious power management and sensor scheduling can effectively extend network lifetime. In this paper we consider the problem of scheduling multimedia sensor activities to maximize network lifetime. The environment is divided in domains monitored by clusters of multimedia sensor nodes. Network lifetime increment is achieved by cooperation between multimedia sensors in two priority-based ways: Intra-cluster cooperation and Inter-cluster cooperation. We will see that the lifetime of cluster nodes is considerably increased under the proposed node selection and scheduling procedures. As for big clusters, the lifetime even is prolonged to 5.5 times with respect to the ordinary un-cooperative node awakening.