Searcharxiv⌕ Search

arXiv subjects

B. Sundar Rajan

Publications and source records attributed to B. Sundar Rajan.

At least 73 records · Page 4Linked to original sources

Secretive Coded Caching with Shared Caches

We consider the problem of \emph{secretive coded caching} in a shared cache setup where the number of users accessing a particular \emph{helper cache} is more than one, and every user can access exactly one helper cache. In secretive coded caching, the constraint of \emph{perfect secrecy} must be satisfied. It requires that the users should not gain, either from their caches or from the transmissions, any information about the content of the files that they did not request from the server. In order to accommodate the secrecy constraint, our problem setup requires, in addition to a helper cache, a dedicated \emph{user cache} of minimum capacity of 1 unit to every user. This is where our formulation differs from the original work on shared caches (``Fundamental Limits of Coded Caching With Multiple Antennas, Shared Caches and Uncoded Prefetching'' by E.~Parrinello, A.~{Ü}nsal and P.~Elia in Trans. Inf. Theory, 2020). In this work, we propose a secretively achievable coded caching scheme with shared caches under centralized placement. When our scheme is applied to the dedicated cache setting, it matches the scheme by Ravindrakumar \emph{et al.} (``Private Coded Caching'', in Trans. Inf. Forensics and Security, 2018).

cs.IT↗

Coded Caching with Shared Caches from Generalized Placement Delivery Arrays

We consider the coded caching problem with shared caches where several users share a cache, but each user has access to only a single cache. For this network, the fundamental limits of coded caching are known for centralized and decentralized settings under uncoded placement. In the centralized case, to achieve the gains offered by coded caching, one requires a sub-packetization which increases exponentially with the number of caches. The dedicated cache networks had a similar issue, and placement delivery arrays (PDAs) were introduced as a solution to it. Using the PDA framework, we propose a procedure to obtain new coded caching schemes for shared caches with lower sub-packetization requirements. The advantage of this procedure is that we can transform all the existing PDA structures into coded caching schemes for shared caches, thus resulting in low sub-packetization schemes. We also show that the optimal scheme given by Parrinello, Unsal and Elia (Fundamental Limits of Coded Caching with Multiple Antennas, Shared Caches and Uncoded Prefetching) can be recovered using a Maddah-Ali Niesen PDA.

cs.IT↗

Secretive Coded Caching from PDAs

The coded caching problem with secrecy constraint i.e., the users should not be able to gain any information about the content of the files that they did not demand, is known as the secretive coded caching problem. This was proposed by Ravindrakumar et al. in the paper titled ``Private Coded Caching'' that appeared in \emph{ IEEE Transactions on Information Forensics and Security}, 2018 and is characterised by subpacketization levels growing exponentially with the number of users. In the context of coded caching without secrecy, coded caching schemes at subexponential subpacketization levels are feasible by representing the caching system in the form of a Placement Delivery Array (PDA) and designing placement and delivery policies from it. Motivated by this, we propose a secretive coded caching scheme with low subpacketization using PDA, for users with dedicated caches in the centralized setting. When our scheme is applied to a special class of PDA known as MN PDA, the scheme proposed by Ravindrakumar et al. is recovered.

cs.IT↗

Improved Multi-access Coded Caching Schemes from Cross Resolvable Designs

Recently multi-access coded caching schemes with number of users different from the number of caches obtained from a special case of resolvable designs called Cross Resolvable Designs (CRDs) have been reported and a new performance metric called rate-per-user has been introduced \cite{KNRarXiv}. In this paper we present a generalization of this work resulting in multi-access coded caching schemes with improved rate-per-user.

cs.IT↗

Maddah-Ali-Niesen Scheme for Multi-access Coded Caching

The well known Maddah-Ali-Niesen (MAN) coded caching scheme for users with dedicated cache is extended for use in multi-access coded cache scheme where the number of users need not be same as the number of caches in the system. The well known MAN scheme is recoverable as a special case of the multi-access system considered. The performance of this scheme is compared with the existing works on multi-access coded caching. To be able to compare the performance of different multi-access schemes with different number of users for the same number of caches, the terminology of per user rate (rate divided by the number of users) introduced in \cite{KNS} is used.

cs.IT↗

Multi-Access Coded Caching with Secure Delivery

The multi-access variant of the coded caching problem in the presence of an external wiretapper is investigated . A multi-access coded caching scheme with $K$ users, $K$ caches and $N$ files, where each user has access to $L$ neighbouring caches in a cyclic wrap-around manner, is proposed, which is secure against the wiretappers. Each transmission in the conventional insecure scheme will be now encrypted by a random key. The proposed scheme uses a novel technique for the key placement in the caches. It is also shown that the proposed secure multi-access coded caching scheme is within a constant multiplicative factor from the information-theoretic optimal rate for $L\geq \frac{K}{2}$ and $N\geq 2K$.

cs.IT↗

Multi-access Coded Caching Scheme with Linear Sub-packetization using PDAs

We consider multi-access coded caching problem introduced by Hachem et.al., where each user has access to $L$ neighboring caches in a cyclic wrap-around fashion. We focus on the deterministic schemes for a specific class of multi-access coded caching problem based on the concept of PDA. We construct new PDAs which specify the delivery scheme for the specific class of multi-access coded caching problem discussed in this paper. For the proposed scheme, the coding gain is larger than that of the state-of-the-art while the sub-packetization level varies only linearly with the number of users. Hence, we achieve a lower transmission rate with the least sub-packetization level compared to the existing schemes.

cs.IT↗

An Improved Multi-access Coded Caching with Uncoded Placement

In this work, we consider a slight variant of well known coded caching problem, referred as multi-access coded caching problem, where each user has access to $z$ neighboring caches in a cyclic wrap-around way. We present a placement and delivery scheme for this problem, under the restriction of uncoded placement. Our work is a generalization of one of the cases considered in "Multi-access coded caching : gains beyond cache-redundancy" by B. Serbetci, E. Parrinello and P. Elia. To be precise, when our scheme is specialized to $z=\frac{K-1}{K γ}$, for any $K γ$, where $K$ is the number of users and $γ$ is the normalized cache size, we show that our result coincides with their result. We show that for the cases considered in this work, our scheme outperforms the scheme proposed in "Rate-memory trade-off for multi-access coded caching with uncoded placement" by K. S. Reddy and N. Karamchandani, except for some special cases considered in that paper. We also show that for $z= K-1$, our scheme achieves the optimal transmission rate.

cs.IT↗

Optimal Demand Private Coded Caching for Users with Small Buffers

Coded Caching is an efficient technique to reduce peak hour network traffic. One limitation of known coded caching schemes is that the demands of all users are revealed to their peers in the delivery phase. Schemes that assure privacy for user demands are studied in recent past. Assuming that the users are equipped with caches of small memory sizes, the achievable rate under demand privacy constraints is investigated in this work. We present an MDS code based demand private coded caching scheme with $K$ users and $N$ files that achieves a memory rate pair $\left(\frac{1}{K(N-1)+1},N\left(1-\frac{1}{K(N-1)+1}\right)\right)$. The presented memory-rate pair meets the lower bound under demand-privacy requirements, proposed by Yan \textit{et al.} in the recent work \cite{c13}. By memory sharing this characterizes the exact rate-memory trade-off for the demand private coded caching scheme for cache memory $M\in \left[0,\frac{1}{K(N-1)+1}\right]$.

cs.IT↗

Multi-access Coded Caching from a New Class of Cross Resolvable Designs

Multi-access coded caching schemes from cross resolvable designs (CRD) have been reported recently \cite{KNRarXiv}. To be able to compare coded caching schemes with different number of users and possibly with different number of caches a new metric called rate-per-user was introduced and it was shown that under this new metric the schemes from CRDs perform better than the Maddah-Ali-Niesen scheme in the large memory regime. In this paper a new class of CRDs is presented and it is shown that the multi-access coded caching schemes derived from these CRDs perform better than the Maddah-Ali-Niesen scheme in the entire memory regime. Comparison with other known multi-access coding schemes is also presented.

cs.IT↗

Decentralized and Online Coded Caching with Shared Caches: Fundamental Limits with Uncoded Prefetching

Decentralized coded caching scheme, introduced by Maddah-Ali and Niesen, assumes that the caches are filled with no coordination. This work identifies a decentralized coded caching scheme -- under the assumption of uncoded placement -- for shared cache network, where each cache serves multiple users. Each user has access to only a single cache and the number of caches is less than or equal to the number of users. For this setting, we derive the optimal worst-case delivery time for any user-to-cache association profile where each such profile describes the number of users served by each cache. The optimality is shown using an index-coding based converse. Further, we improve the delivery scheme to accommodate redundant demands. Also, an optimal linear error correcting delivery scheme is proposed for the worst-case demand scenario. Next, we consider the Least Recently Sent (LRS) online coded caching scheme where the caches need to be updated based on the sequence of demands made by the users. Cache update happens if any of the demanded file was not partially cached at the users. The update is done by replacing the least recently sent file with the new file. But, the least recently sent file need not be unique. In that case, there needs to be some ordering of the files which are getting partially cached, or else centralized coordination would have to be assumed which does not exist. If each user removes any of the least recently used files at random, then the next delivery phase will not serve the purpose. A modification is suggested for the scheme by incorporating an ordering of files. Moreover, all the above results with shared caches are extended to the online setting.

cs.IT↗

An Embedded Index Code Construction Using Sub-packetization

A variant of the index coding problem (ICP), the embedded index coding problem (EICP) was introduced in [A. Porter and M. Wootters, "Embedded index coding," ITW, Sweden, 2019] which was motivated by its application in distributed computing where every user can act as sender for other users and an algorithm for code construction was reported. The constructions depends on the computation of minrank of a matrix, which is computationally intensive. In [A. Mahesh, N. Sageer Karat and B. S. Rajan, "Min-rank of Embedded Index Coding Problems," ISIT, 2020], for EICP, a notion of side-information matrix was introduced and it was proved that the length of an optimal scalar linear index code is equal to the min-rank of the side-information matrix. The authors have provided an explicit code construction for a class of EICP - \textit{Consecutive and Symmetric Embedded Index Coding Problem (CS-EICP)}. We introduce the idea of sub-packetization of the messages in index coding problems to provide a novel code construction for CS-EICP in contrast to the scalar linear solutions provided in the prior works. For CS-EICP, the normalized rate, which is defined as the number of bits transmitted by all the users together normalized by the total number of bits of all the messages, for our construction is lesser than the normalized rate achieved by Mahesh et al.,for scalar linear codes.

cs.IT↗

A Coded Caching Scheme with Linear Sub-packetization and its Application to Multi-Access Coded Caching

This paper addresses the problem of exponentially increasing sub-packetization with the number of users in a centralized coded caching system by introducing a new coded caching scheme inspired by the symmetric neighboring consecutive side information index coding problem. The scheme has a placement policy where the number of sub-packets required grows only linearly with the number of users, with no restriction on file size, and a delivery policy which is instantaneously decodable. Further, an application of the new delivery scheme in a multi-access coded caching set-up is studied and a few results in that direction are presented. In particular, in the multi-access set-up, for cases where optimality rate-memory trade-off characterizations are available, it is shown that the new delivery scheme achieves optimal or near-optimal rates.

cs.IT↗

Index Coding in Vehicle to Vehicle Communication

Vehicle to Vehicle (V2V) communication phase is an integral part of collaborative message dissemination in vehicular ad-hoc networks (VANETs). In this work, we apply index coding techniques to reduce the number of transmissions required for data exchange. The index coding problem has a sender, which tries to meet the demands of several receivers in a minimum number of transmissions. All these receivers have some prior knowledge of the messages, known as the side-information. In this work, we consider a particular case of the index coding problem, where multiple nodes want to share information among them. Under this set up, lower bound on the number of transmissions is established when the cardinality of side-information is the same. An optimal solution to achieve the bound in a special case of VANET scenario is presented. For this special case, we consider the link between the nodes to be error-prone, and in this setting, we construct optimal linear error correcting index codes.

cs.IT↗

Multi-access Coded Caching Schemes From Cross Resolvable Designs

We present a novel caching and coded delivery scheme for a multi-access network where multiple users can have access to the same cache (shared cache) and any cache can assist multiple users. This scheme is obtained from resolvable designs satisfying certain conditions which we call {\it cross resolvable designs}. To be able to compare different multi-access coded schemes with different number of users we normalize the rate of the schemes by the number of users served. Based on this per-user-rate we show that our scheme performs better than the well known Maddah-Ali - Nieson (MaN) scheme and the recently proposed ("Multi-access coded caching : gains beyond cache-redundancy" by Serbetci, Parrinello and Elia) SPE scheme. It is shown that the resolvable designs from affine planes are cross resolvable designs and our scheme based on these performs better than the MaN scheme for large memory size cases. The exact size beyond which our performance is better is also presented. The SPE scheme considers only the cases where the product of the number of users and the normalized cache size is 2, whereas the proposed scheme allows different choices depending on the choice of the cross resolvable design.

cs.IT↗

Straggler Mitigation with Tiered Gradient Codes

Coding theoretic techniques have been proposed for synchronous Gradient Descent (GD) on multiple servers to mitigate stragglers. These techniques provide the flexibility that the job is complete when any $k$ out of $n$ servers finish their assigned tasks. The task size on each server is found based on the values of $k$ and $n$. However, it is assumed that all the $n$ jobs are started when the job is requested. In contrast, we assume a tiered system, where we start with $n_1\ge k$ tasks, and on completion of $c$ tasks, we start $n_2-n_1$ more tasks. The aim is that as long as $k$ servers can execute their tasks, the job gets completed. This paper exploits the flexibility that not all servers are started at the request time to obtain the achievable task sizes on each server. The task sizes are in general lower than starting all $n_2$ tasks at the request times thus helping achieve lower task sizes which helps to reduce both the job completion time and the total server utilization.

cs.IT↗

Optimal Scalar Linear Codes for Some Classes of The Two-Sender Groupcast Index Coding Problem

The two-sender groupcast index coding problem (TGICP) consists of a set of receivers, where all the messages demanded by the set of receivers are distributed among the two senders. The senders can possibly have a set of messages in common. Each message can be demanded by more than one receiver. Each receiver has a subset of messages (known as its side information) and demands a message it does not have. The objective is to design scalar linear codes at the senders with the minimum aggregate code length such that all the receivers are able to decode their demands, by leveraging the knowledge of the side information of all the receivers. In this work, optimal scalar linear codes of three sub-problems (considered as single-sender groupcast index coding problems (SGICPs)) of the TGICP are used to construct optimal scalar linear codes for some classes of the TGICP. We introduce the notion of joint extensions of a finite number of SGICPs, which generalizes the notion of extensions of a single SGICP introduced in a prior work. An SGICP $\mathcal{I}_E$ is said to be a joint extension of a finite number of SGICPs if all the SGICPs are disjoint sub-problems of $\mathcal{I}_E$. We identify a class of joint extensions, where optimal scalar linear codes of the joint extensions can be constructed using those of the sub-problems. We then construct scalar linear codes for some classes of the TGICP, when one or more sub-problems of the TGICP belong to the above identified class of joint extensions, and provide some necessary conditions for the optimality of the construction.

cs.IT↗

Optimal Weakly Secure Linear Codes for Some Classes of the Two-Sender Index Coding Problem

The two-sender unicast index coding problem is the most fundamental multi-sender index coding problem. The two senders collectively cater to the demands of all the receivers, by taking advantage of the knowledge of their side-information. Each receiver demands a unique message and has some side-information. Weakly secure index coding problem is a practical version of the index coding problem in the presence of an eavesdropper. The eavesdropper can not gain any information about the messages he does not have, by listening to the senders' transmissions. We provide constructions of weakly secure linear codes for different classes of the two-sender unicast index coding problem, using those of its sub-problems. The constructions are valid only if such codes exist for all the sub-problems under consideration. We identify some classes of the two-sender problem, where the constructions provide optimal weakly secure linear index codes.

cs.IT↗