With the increasing adoption of Convolutional Neural Networks (CNNs) in edge and embedded applications, there is a growing demand for hardware accelerators that provide high computational accuracy while maintaining low energy consumption. In this work, we present a reconfigurable CNN accelerator specifically designed for real-time object detection. The accelerator is based on a systolic array architecture and integrates the posit number system to enhance both accuracy and dynamic range in numerical representation. To further improve energy efficiency, the design incorporates a set of lightweight approximation techniques applied to arithmetic operations. In particular, the Multiply and Accumulate (MAC) units employ approximate strategies for multiplication, normalization, and rounding of the fraction part, selectively simplifying low-significance computations to reduce hardware cost, while ensuring that the final classification accuracy remains within acceptable limits. The proposed architecture is implemented at the Register-Transfer Level (RTL) level and evaluated on a Xilinx ZCU102 FPGA using ResNet-50 and the ImageNet dataset. Experimental results demonstrate a 21% reduction in power consumption and an 18% decrease in resource utilization compared to conventional posit-based implementations. The proposed 32 32 systolic array achieves a 3% improvement in inference accuracy compared to the floating-point implementation. These findings indicate that the combination of posit arithmetic with targeted approximation techniques provides a promising solution for achieving accurate and energy-efficient deep learning inference at the edge.

Abstract Growing demands for pervasive and ubiquitous services over wireless mobile networks and evolution of such networks towards heterogeneous solutions have emphasized the necessity of more intelligent handoff decisions. The existing handoff management methods in the literature are mostly using signal strength measurements and other link quality evaluations not addressing the knowledge about context of mobile devices, users and networks. Recently, context‐aware handoff management has been considered as a novel candidate for fourth generation (4G) wireless technology. In this paper, user perceived quality of service has been considered in addition to traditional contexts such as user preferences, application requirements, network parameters and link quality for decision making. User perceived quality (UPQ) has been employed as a trigger source, in addition to link layer triggers which are emerged using media independent handover (MIH) event service. This paper presents a policy based mechanism for handoff decision making where fuzzy petri nets (FPNs) have been utilized as its evaluation algorithm. A case study has been provided by simulations to show the usability and user level satisfaction. Simulation results show superior performance in terms of UPQ, jitter and packet delivery measures. Copyright © 2009 John Wiley & Sons, Ltd.

Stream ciphering is an essential coding method for delay sensitive traffic such as voice and video. In this method the source sequence is combined with the output of a key stream generator using XOR The strength and robustness of this ciphering method lays in complexity of the key stream generator. A well-known method for generating key stream is pseudo-random number generator) which is based on LFSR. Recently the anti-random sequence has attained attraction and shown to have better statistical properties compared to pseudo-random numbers, so it appears be suitable for stream ciphering. The existing methods for generating anti-random numbers are highly time and memory consuming. This paper presents an approximate method for generating anti-random sequence, which is very simple and does not need large capacity of memory, while preserves the properties of the sequence. Moreover, we show that our method presents enough statistical complexity to fulfill the required security for stream ciphering.