Minimizing Latency for Secure Coded Computing Using Secret Sharing via Staircase Codes

被引:34
作者
Bitar, Rawad [1 ]
Parag, Parimal [2 ]
El Rouayheb, Salim [1 ]
机构
[1] Rutgers State Univ, ECE Dept, Piscataway, NJ 08854 USA
[2] Indian Inst Sci, Dept Elect Commun Engn, Bengaluru 560012, India
关键词
Cryptography; Delays; Decoding; Distributed computing; Privacy; Encoding; Task analysis; data privacy; secret sharing; secure coded computing; machine learning;
D O I
10.1109/TCOMM.2020.2988506
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
We consider the setting of a Master server, M, who possesses confidential data and wants to run intensive computations on it, as part of a machine learning algorithm for example. The Master wants to distribute these computations to untrusted workers who volunteered to help with this task. However, the data must be kept private in an information theoretic sense. Some of the workers may be stragglers, e.g., slow or busy. We are interested in reducing the delays experienced by the Master. We focus on linear computations as an essential operation in many iterative algorithms. We propose a solution based on new codes, called Staircase codes, introduced previously by two of the authors. Staircase codes allow flexibility in the number of stragglers up to a given maximum, and universally achieve the information theoretic limit on the download cost by the Master, leading to latency reduction. We find upper and lower bounds on the Master's mean waiting time. We derive the distribution of the Master's waiting time, and its mean, for systems with up to two stragglers. We show that Staircase codes always outperform existing solutions based on classical secret sharing codes. We validate our results with extensive implementation on Amazon EC2.
引用
收藏
页码:4609 / 4619
页数:11
相关论文
共 42 条
[1]  
[Anonymous], 2016, IEEE INT SYMP INFO
[2]  
[Anonymous], 2019, IEEE INT SYMP INFO
[3]  
Atallah M. J., 2010, P 5 ACM S INF COMP C, P48
[4]  
Bitar R., 2018, ARXIV180202640
[5]   Staircase Codes for Secret Sharing With Optimal Communication and Read Overheads [J].
Bitar, Rawad ;
El Rouayheb, Salim .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2018, 64 (02) :933-943
[6]  
Bitar R, 2017, IEEE INT SYMP INFO, P2900, DOI 10.1109/ISIT.2017.8007060
[7]  
Bitar R, 2016, IEEE INT SYMP INFO, P1396, DOI 10.1109/ISIT.2016.7541528
[8]   EFFICIENT FULLY HOMOMORPHIC ENCRYPTION FROM (STANDARD) LWE [J].
Brakerski, Zvika ;
Vaikuntanathan, Vinod .
SIAM JOURNAL ON COMPUTING, 2014, 43 (02) :831-871
[9]   On the Upload versus Download Cost for Secure and Private Matrix Multiplication [J].
Chang, Wei-Ting ;
Tandon, Ravi .
2019 IEEE INFORMATION THEORY WORKSHOP (ITW), 2019, :469-473
[10]  
Chang WT, 2018, IEEE GLOB COMM CONF