FIMPA: A Fixed Identity Mapping Prediction Algorithm in Edge Computing Environment

Abstract

Edge computing is a research hotspot that extends cloud computing to the edge of the network. Due to the recent developments in computation, storage and network technology for end devices, edge networks have become more powerful, making it possible to integrate locator/identity separation protocol (LISP) into these networks. Accordingly, in this paper, we introduce LISP into edge routers at the edge network, focusing primarily on the delay problem of mapping resolution and cache updating in LISP with the help of edge computing. To solve this delay problem, we first analyze the communication process of the locator/identity separation network and consider using the prediction method to underpin this research. In order to achieve a good prediction result, we propose and implement a Fixed Identity Mapping Prediction Algorithm (FIMPA) based on collaborative filtering, and further verify the effectiveness of the proposed algorithm through experiments on real-world data.

Energy-Efficient Model Compression and Splitting for Collaborative Inference Over Time-Varying Channels

“…In fact, the cut layer index belongs to a discrete set, which makes problem (8) a combinatorial one. Next, we propose an approximate and a low-complexity approach to solve problem (8).…”

“…Since the loss function does not depend on the choice of the cut layer (transmitted representation), problem (7) can be re-written as (8) where θ = [θ c,i , θ s,i ] is the whole model, which is constant across different transmissions.…”

“…However, the large size of the data to be shared as well as the privacy constraint limit the feasibility of this approach in practice [6]. Recently, embedded devices equipped with more enhanced computation capabilities allow for more effective data processing at the network edge [7], [8]. However, for large-size models and under edge devices' energy constraints, the inference task may not be locally performed.…”

Adaptive weighted dynamic differential evolution algorithm for emergency material allocation and scheduling

“…[8][9][10][11] It is difficult to solve the problem using traditional mathematical methods such as integer programming and divide-and-conquer algorithms. In recent years, genetic algorithms, particle swarm optimization, and other heuristic optimization algorithms have been widely used to solve the problem of emergency material allocation and scheduling, [12][13][14][15][16][17][18] but these algorithms generally have long search time, easy stagnation, and easy local optimal solutions, resulting in a long solution time for emergency material allocation scheduling. Therefore, it is the key point of this article to construct an emergency material allocation and scheduling model based on the actual problem characteristics of emergencies, and then find an efficient algorithm suitable for the model.…”

Adaptive Dynamic Disturbance Strategy for Differential Evolution Algorithm

“…As a highly efficient parallel search algorithm with few controlled parameters, the DE algorithm has been widely used in neuron networks [3,4], power systems [5,6], vehicle routing problems [7][8][9][10][11] and many other fields [12][13][14][15][16][17][18][19][20]. Like other intelligent algorithms, however, the DE algorithm also has disadvantages such as precocity, strong parameter dependence and difficulty in obtaining global optimum values for high-dimensional complex objective functions [21][22][23][24][25][26][27]. Therefore, researchers proposed a number of improvement strategies for DE's existing shortcomings.…”