Multi-access Edge Computing (MEC) is an emerging computing architecture to release the resource burden of the centralized cloud and reduce the mobile application latency. Services management and MEC requests routing i...
详细信息
Multi-access Edge Computing (MEC) is an emerging computing architecture to release the resource burden of the centralized cloud and reduce the mobile application latency. Services management and MEC requests routing is a major problem in MEC systems. Existing works mainly focus on the one-hop centralized request routing strategies. However, the centralized one-hop routing method is not suitable enough since the MEC network is a distributed system, and the number of MEC requests increases dramatically. In this paper, we have proposed an online problem. In such problem, we jointly consider the mobile edge service management and the distributed multi-hop requests routing in an MEC network in which the MEC requests randomly generate. We prove that such problem is NP-Hard even in the off-line scenario. Furthermore, we propose an approximation algorithm to manage the MEC services and two distributed online algorithms to route MEC requests. The approximation ratio and competitive ratio of these algorithms have been analyzed. Experiments are carried out to evaluate the performance of the algorithms and simulation results imply that these algorithms are effective and efficient.
Network management protocols often require timely and meaningful insight about per flow network traffic. This paper introduces Randomized Admission Policy (RAP) -a novel algorithm for the frequency, top-k, and byte vo...
详细信息
Network management protocols often require timely and meaningful insight about per flow network traffic. This paper introduces Randomized Admission Policy (RAP) -a novel algorithm for the frequency, top-k, and byte volume estimation problems, which are fundamental in network monitoring. We demonstrate space reductions compared to the alternatives, for the frequency estimation problem, by a factor of up to 32 on real packet traces and up to 128 on heavy-tailed workloads. For top-k identification, RAP exhibits memory savings by a factor of between 4 and 64 depending on the workloads' skewness. These empirical results are backed by formal analysis, indicating the asymptotic space improvement of our probabilistic admission approach. In Addition, we present d-way RAP, a hardware friendly variant of RAP that empirically maintains its space and accuracy benefits.
We study a recently introduced path coloring problem with applications to wavelength assignment in all-optical networks with multiple fibers. In contrast to classical path coloring, it is, in this setting, possible to...
详细信息
We study a recently introduced path coloring problem with applications to wavelength assignment in all-optical networks with multiple fibers. In contrast to classical path coloring, it is, in this setting, possible to assign a color more than once to paths that pass through the same edge;the number of allowed repetitions per edge is given and the goal is to minimize the number of colors used. We present algorithms and hardness results for tree topologies of special interest. Our algorithms achieve approximation ratio of 2 in spiders and 3 in caterpillars, whereas the best algorithm for trees so far, achieves an approximation ratio of 4. We also study the directed version of the problem and show that it admits a 3-approximation algorithm in caterpillars, while it can be solved exactly in spiders.
It is not difficult to anticipate that the system capacity and spectral efficiency of 5G will be greater than those of 4G, and so will the number of network-connected wireless devices. Moreover, the infrastructure of ...
详细信息
It is not difficult to anticipate that the system capacity and spectral efficiency of 5G will be greater than those of 4G, and so will the number of network-connected wireless devices. Moreover, the infrastructure of wireless communications today will not meet the requirements of a 5G environment because a large number of traffic flows will be generated between a large number of heterogeneous devices. The deployment strategy is one of the promising solutions to meet the expected demands of 5G. This article begins with a brief review of the deployment problem of wireless communication and a hyper-dense deployment problem (HDDP) definition for the requirements of 5G. A simple example for illustrating how a metaheuristic algorithm can be used to solve the HDDP is then given. Finally, some open and possible research issues are discussed to suggest future research trends on this problem.
Efficient collaboration between collaborative machine learning and wireless communication technology, forming a Federated Edge Learning (FEEL), has spawned a series of next-generation intelligent applications. However...
详细信息
Efficient collaboration between collaborative machine learning and wireless communication technology, forming a Federated Edge Learning (FEEL), has spawned a series of next-generation intelligent applications. However, due to the openness of network connections, the FEEL framework generally involves hundreds of remote devices (or clients), resulting in expensive communication costs, which is not friendly to resource-constrained FEEL. To address this issue, we propose a distributed approximate Newton-type algorithm with fast convergence speed to alleviate the problem of FEEL resource (in terms of communication resources) constraints. Specifically, the proposed algorithm is improved based on distributed L-BFGS algorithm and allows each client to approximate the high-cost Hessian matrix by computing the low-cost Fisher matrix in a distributed manner to find a "better" descent direction, thereby speeding up convergence. Second, we prove that the proposed algorithm has linear convergence in strongly convex and non-convex cases and analyze its computational and communication complexity. Similarly, due to the heterogeneity of the connected remote devices, FEEL faces the challenge of heterogeneous data and non-IID (Independent and Identically Distributed) data. To this end, we design a simple but elegant training scheme, namely FedOVA (Federated One-vs-All), to solve the heterogeneous statistical challenge brought by heterogeneous data. In this way, FedOVA first decomposes a multi-class classification problem into more straightforward binary classification problems and then combines their respective outputs using ensemble learning. In particular, the scheme can be well integrated with our communication efficient algorithm to serve FEEL. Numerical results verify the effectiveness and superiority of the proposed algorithm.
Cloud radio access network (C-RAN) is a promising 5G network architecture by establishing baseband units (BBU) pools to perform baseband processing functionalities and deploying remote radio heads (RRHs) for wireless ...
详细信息
Cloud radio access network (C-RAN) is a promising 5G network architecture by establishing baseband units (BBU) pools to perform baseband processing functionalities and deploying remote radio heads (RRHs) for wireless signal transmission and reception. Mobile-edge computing (MEC) offers a way to shorten the service delay by building small-scale cloud infrastructures at the network edge. By co-locating the BBU pool with edge cloud at the so-called BBU node, we can take full advantages of C-RAN and MEC for better spectrum utilization and delay-guaranteed services. In this article, we first study how to allocate each users task to the BBU node and find the path from his/her accessing RRH node to the BBU node such that the maximum service delay among all the requests is minimized. We then consider this problem with survivability concerns, which is to use both primary and backup BBU nodes to issue the request such that the primary path and backup path are link disjoint. We analyze the complexities of these two problems and prove they are NP-hard in general. Subsequently, we devise a randomized approximation algorithm and an efficient heuristic to solve the considered problems, respectively. The simulation results show that the proposed algorithms outperform two benchmark heuristics in terms of acceptance ratio and maximum service delay.
Clustering problems in a complex geographical setting are often required to incorporate the type and extent of land cover within a region. Given a set P of n points in a geographical setting. with the constraint that ...
详细信息
Clustering problems in a complex geographical setting are often required to incorporate the type and extent of land cover within a region. Given a set P of n points in a geographical setting. with the constraint that the points of P can only occur in one type of land cover, an interesting problem is the detection Of Clusters. First, we extend the definition of clusters and define the concept of a region-restricted cluster that satisfies the following properties: (i) the cluster has sufficient number of points, (ii) the cluster points are confined to a small geographical area, and (iii) the amount of land cover of the specific type in which the points lie is also small. Next, we give efficient exact and approximation algorithms for computing such clusters. The exact algorithm determines all axis-parallel squares with exactly m out of n points inside, size at most some prespecified value, and area of a given land cover type at most another prespecified value, and runs in O(nm log(2) n + (nm + nn(f)) log(2) n(f)) time, where nf is the number of edges that bound the regions with the given land cover type. The approximation algorithm allows the square to be a factor 1 + epsilon too large, and runs in O (n log n + n/epsilon(2) + n(f) log(2) n(f) + (n log(2) n(f))/(m epsilon(2))) time. We also show how to compute largest clusters and outliers. Crown Copyright (c) 2008 Published by Elsevier B.V. All rights reserved.
We show the existence of a fully polynomial-time approximation scheme (FPTAS) for the problem of maximizing a non-negative polynomial over mixed-integer sets in convex polytopes, when the number of variables is fixed....
详细信息
We show the existence of a fully polynomial-time approximation scheme (FPTAS) for the problem of maximizing a non-negative polynomial over mixed-integer sets in convex polytopes, when the number of variables is fixed. Moreover, using a weaker notion of approximation, we show the existence of a fully polynomial-time approximation scheme for the problem of maximizing or minimizing an arbitrary polynomial over mixed-integer sets in convex polytopes, when the number of variables is fixed.
Given dissimilarity data on pairs of objects in a set, we study the problem of fitting a tree metric to this data so as to minimize additive error (i.e., some measure of the difference between the tree metric and the ...
详细信息
Given dissimilarity data on pairs of objects in a set, we study the problem of fitting a tree metric to this data so as to minimize additive error (i.e., some measure of the difference between the tree metric and the given data). This problem arises in constructing an M-level hierarchical clustering of objects (or an ultrametric on objects) so as to match the given dissimilarity data-a basic problem in statistics. Viewed in this way, the problem is a generalization of the correlation clustering problem (which corresponds to M - 1). We give a very simple randomized combinatorial algorithm for the M-level hierarchical clustering problem that achieves an approximation ratio of M+2. This is a generalization of a previous factor 3 algorithm for correlation clustering on complete graphs. The problem of fitting tree metrics also arises in phylogeny where the objective is to learn the evolution tree by fitting a tree to dissimilarity data on taxa. The quality of the fit is measured by taking the l(p) norm of the difference between the tree metric constructed and the given data. Previous results obtained a factor 3 approximation for finding the closest tree metric under the l(infinity) norm. No nontrivial approximation for general l(p) norms was known before. We present a novel linear program formulation for this problem and obtain an O((log n log log n)(1/p))-approximation to the closest ultrametric under the l(p) norm using this. Our techniques are based on representing and viewing an ultrametric as a hierarchy of clusterings and may be useful in other contexts.
We consider the generalized minimum Manhattan network problem (GMMN). The input to this problem is a set R of n pairs of terminals, which are points in . The goal is to find a minimum-length rectilinear network that c...
详细信息
We consider the generalized minimum Manhattan network problem (GMMN). The input to this problem is a set R of n pairs of terminals, which are points in . The goal is to find a minimum-length rectilinear network that connects every pair in R by a Manhattan path, that is, a path of axis-parallel line segments whose total length equals the pair's Manhattan distance. This problem is a natural generalization of the extensively studied minimum Manhattan network problem (MMN) in which R consists of all possible pairs of terminals. Another important special case is the well-known rectilinear Steiner arborescence problem (RSA). As a generalization of these problems, GMMN is NP-hard. No approximation algorithms are known for general GMMN. We obtain an -approximation algorithm for GMMN. Our solution is based on a stabbing technique, a novel way of attacking Manhattan network problems. Some parts of our algorithm generalize to higher dimensions, yielding a simple -approximation algorithm for the problem in arbitrary fixed dimension d. As a corollary, we obtain an exponential improvement upon the previously best -ratio for MMN in d dimensions (ESA 2011). En route, we show that an existing -approximation algorithm for 2D-RSA generalizes to higher dimensions.
暂无评论