ON THE COMMUNICATION COMPLEXITY OF DISTRIBUTED ALGEBRAIC COMPUTATION

被引:7
|
作者
LUO, ZQ [1 ]
TSITSIKLIS, JN [1 ]
机构
[1] MIT,CAMBRIDGE,MA 02139
关键词
ALGORITHMS; THEORY; ALGEBRAIC COMPUTATION; COMMUNICATION COMPLEXITY; LOWER BOUND;
D O I
10.1145/174147.174149
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
We consider a situation where two processors P1 and P2 arc to evaluate a collection of functions f1,..., f(s) of two-vector variables x, y, under the assumption that processor P1 (respectively, P2) has access only to the value of the variable x (respectively, y) and the functional form of f1,..., f(s). We provide some new bounds on the communication complexity (the amount of information that has to be exchanged between the processors) for this problem. An almost optimal bound is derived for the case of one-way communication when the functions f1,..., f(s) are polynomials. We also derive some new lower hounds for the case of two-way communication that improve on earlier bounds by Abelson [2]. As an application, we consider the case where x and y are n X n matrices and f(x, y) is a particular entry of the inverse of x + y. Under a certain restriction on the class of allowed communication protocols, we obtain an OMEGA(n2) lower bound, in contrast to the OMEGA(n) lower bound obtained by applying Abelson's results. Our results are based on certain tools from classical algebraic geometry and field extension theory.
引用
收藏
页码:1019 / 1047
页数:29
相关论文
共 50 条
  • [41] Distributed Function Computation in Asymmetric Communication Scenarios
    Agnihotri, Samar
    Venkatachalapathy, Rajesh
    Jamadagni, H. S.
    2009 47TH ANNUAL ALLERTON CONFERENCE ON COMMUNICATION, CONTROL, AND COMPUTING, VOLS 1 AND 2, 2009, : 724 - +
  • [42] On the Tradeoff Between Computation and Communication Costs for Distributed Linearly Separable Computation
    Wan, Kai
    Sun, Hua
    Ji, Mingyue
    Caire, Giuseppe
    IEEE TRANSACTIONS ON COMMUNICATIONS, 2021, 69 (11) : 7390 - 7405
  • [43] COMPLEXITY OF THE COMPUTATION OF THE CANONICAL WHITNEY STRATIFICATION OF AN ALGEBRAIC SET IN C(N)
    MOSTOWSKI, T
    RANNOU, E
    LECTURE NOTES IN COMPUTER SCIENCE, 1991, 539 : 281 - 291
  • [44] On the communication complexity of Lipschitzian optimization for the coordinated model of computation
    Mesbahi, M
    JOURNAL OF COMPLEXITY, 2000, 16 (02) : 459 - 473
  • [45] THE COMPUTATION AND COMMUNICATION COMPLEXITY OF A PARALLEL BANDED SYSTEM SOLVER
    LAWRIE, DH
    SAMEH, AH
    ACM TRANSACTIONS ON MATHEMATICAL SOFTWARE, 1984, 10 (02): : 185 - 195
  • [46] Communication Complexity of Approximate Matching in Distributed Graphs
    Huang, Zengfeng
    Radunovic, Bozidar
    Vojnovic, Milan
    Zhang, Qin
    32ND INTERNATIONAL SYMPOSIUM ON THEORETICAL ASPECTS OF COMPUTER SCIENCE (STACS 2015), 2015, 30 : 460 - 473
  • [47] ON THE AVERAGE COMMUNICATION COMPLEXITY OF ASYNCHRONOUS DISTRIBUTED ALGORITHMS
    TSITSIKLIS, JN
    STAMOULIS, GD
    JOURNAL OF THE ASSOCIATION FOR COMPUTING MACHINERY, 1995, 42 (02): : 382 - 400
  • [48] Distributed communication complexity of spanning tree construction
    Vyalyi, M. N.
    Khuziev, I. M.
    PROBLEMS OF INFORMATION TRANSMISSION, 2015, 51 (01) : 49 - 65
  • [49] Communication Complexity of Distributed Convex Learning and Optimization
    Arjevani, Yossi
    Shamir, Ohad
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 28 (NIPS 2015), 2015, 28
  • [50] On the average communication complexity of asynchronous distributed algorithms
    Massachusetts Inst of Technology, Cambridge, United States
    Journal of the ACM, 1995, 42 (02): : 382 - 400