Searcharxiv⌕ Search

arXiv subjects

Jaikumar Radhakrishnan

Publications and source records attributed to Jaikumar Radhakrishnan.

32 records · Page 2Linked to original sources
arXiv2015

How Hard is Computing Parity with Noisy Communications?

We show a tight lower bound of $Ω(N \log\log N)$ on the number of transmissions required to compute the parity of $N$ input bits with constant error in a noisy communication network of $N$ randomly placed sensors, each having one input bit and communicating with others using local transmissions with power near the connectivity threshold. This result settles the lower bound question left open by Ying, Srikant and Dullerud (WiOpt 06), who showed how the sum of all the $N$ bits can be computed using $O(N \log\log N)$ transmissions. The same lower bound has been shown to hold for a host of other functions including majority by Dutta and Radhakrishnan (FOCS 2008). Most works on lower bounds for communication networks considered mostly the full broadcast model without using the fact that the communication in real networks is local, determined by the power of the transmitters. In fact, in full broadcast networks computing parity needs $θ(N)$ transmissions. To obtain our lower bound we employ techniques developed by Goyal, Kindler and Saks (FOCS 05), who showed lower bounds in the full broadcast model by reducing the problem to a model of noisy decision trees. However, in order to capture the limited range of transmissions in real sensor networks, we adapt their definition of noisy decision trees and allow each node of the tree access to only a limited part of the input. Our lower bound is obtained by exploiting special properties of parity computations in such noisy decision trees.

cs.DC↗
arXiv2011

Streaming algorithms for language recognition problems

We study the complexity of the following problems in the streaming model. Membership testing for \DLIN We show that every language in \DLIN\ can be recognised by a randomized one-pass $O(\log n)$ space algorithm with inverse polynomial one-sided error, and by a deterministic p-pass $O(n/p)$ space algorithm. We show that these algorithms are optimal. Membership testing for \LL$(k)$ For languages generated by \LL$(k)$ grammars with a bound of $r$ on the number of nonterminals at any stage in the left-most derivation, we show that membership can be tested by a randomized one-pass $O(r\log n)$ space algorithm with inverse polynomial (in $n$) one-sided error. Membership testing for \DCFL We show that randomized algorithms as efficient as the ones described above for \DLIN\ and $\LL(k)$ (which are subclasses of \DCFL) cannot exist for all of \DCFL: there is a language in \VPL\ (a subclass of \DCFL) for which any randomized p-pass algorithm with error bounded by $ε< 1/2$ must use $Ω(n/p)$ space. Degree sequence problem We study the problem of determining, given a sequence $d_1, d_2,..., d_n$ and a graph $G$, whether the degree sequence of $G$ is precisely $d_1, d_2,..., d_n$. We give a randomized one-pass $O(\log n)$ space algorithm with inverse polynomial one-sided error probability. We show that our algorithms are optimal. Our randomized algorithms are based on the recent work of Magniez et al. \cite{MMN09}; our lower bounds are obtained by considering related communication complexity problems.

cs.DS↗
arXiv2010

An entropy based proof of the Moore bound for irregular graphs

We provide proofs of the following theorems by considering the entropy of random walks: Theorem 1.(Alon, Hoory and Linial) Let G be an undirected simple graph with n vertices, girth g, minimum degree at least 2 and average degree d: Odd girth: If g=2r+1,then n \geq 1 + d*(\Sum_{i=0}^{r-1}(d-1)^i) Even girth: If g=2r,then n \geq 2*(\Sum_{i=0}^{r-1} (d-1)^i) Theorem 2.(Hoory) Let G = (V_L,V_R,E) be a bipartite graph of girth g = 2r, with n_L = |V_L| and n_R = |V_R|, minimum degree at least 2 and the left and right average degrees d_L and d_R. Then, n_L \geq \Sum_{i=0}^{r-1}(d_R-1)^{i/2}(d_L-1)^{i/2} n_R \geq \Sum_{i=0}^{r-1}(d_L-1)^{i/2}(d_R-1)^{i/2}

cs.DM↗
arXiv2008

Optimal Direct Sum and Privacy Trade-off Results for Quantum and Classical Communication Complexity

We show optimal Direct Sum result for the one-way entanglement-assisted quantum communication complexity for any relation f subset of X x Y x Z. We show: Q^{1,pub}(f^m) = Omega(m Q^{1,pub}(f)), where Q^{1,pub}(f), represents the one-way entanglement-assisted quantum communication complexity of f with error at most 1/3 and f^m represents m-copies of f. Similarly for the one-way public-coin classical communication complexity we show: R^{1,pub}(f^m) = Omega(m R^{1,pub}(f)), where R^{1,pub}(f), represents the one-way public-coin classical communication complexity of f with error at most 1/3. We show similar optimal Direct Sum results for the Simultaneous Message Passing quantum and classical models. For two-way protocols we present optimal Privacy Trade-off results leading to a Weak Direct Sum result for such protocols. We show our Direct Sum and Privacy Trade-off results via message compression arguments which also imply a new round elimination lemma in quantum communication. This allows us to extend classical lower bounds on the cell probe complexity of some data structure problems, e.g. Approximate Nearest Neighbor Searching on the Hamming cube {0,1}^n and Predecessor Search to the quantum setting. In a separate result we show that Newman's technique of reducing the number of public-coins in a classical protocol cannot be lifted to the quantum setting. We do this by defining a general notion of black-box reduction of prior entanglement that subsumes Newman's technique. We prove that such a black-box reduction is impossible for quantum protocols. In the final result in the theme of message compression, we provide an upper bound on the problem of Exact Remote State Preparation.

cs.DC↗
arXiv2007

A theorem about relative entropy of quantum states with an application to privacy in quantum communication

We prove the following theorem about relative entropy of quantum states. "Substate theorem: Let rho and sigma be quantum states in the same Hilbert space with relative entropy S(rho|sigma) = Tr rho (log rho - log sigma) = c. Then for all epsilon > 0, there is a state rho' such that the trace distance ||rho' - rho||_t = Tr sqrt{(rho' - rho)^2} <= epsilon, and rho'/2^{O(c/epsilon^2)} <= sigma." It states that if the relative entropy of rho and sigma is small, then there is a state rho' close to rho, i.e. with small trace distance ||rho' - rho||_t, that when scaled down by a factor 2^{O(c)} `sits inside', or becomes a `substate' of, sigma. This result has several applications in quantum communication complexity and cryptography. Using the substate theorem, we derive a privacy trade-off for the set membership problem in the two-party quantum communication model. Here Alice is given a subset A of [n], Bob an input i in [n], and they need to determine if i in A. "Privacy trade-off for set membership: In any two-party quantum communication protocol for the set membership problem, if Bob reveals only k bits of information about his input, then Alice must reveal at least n/2^{O(k)} bits of information about her input." We also discuss relationships between various information theoretic quantities that arise naturally in the context of the substate theorem.

quant-ph↗
arXiv2005

On the Power of Random Bases in Fourier Sampling: Hidden Subgroup Problem in the Heisenberg Group

The hidden subgroup problem (HSP) provides a unified framework to study problems of group-theoretical nature in quantum computing such as order finding and the discrete logarithm problem. While it is known that Fourier sampling provides an efficient solution in the abelian case, not much is known for general non-abelian groups. Recently, some authors raised the question as to whether post-processing the Fourier spectrum by measuring in a random orthonormal basis helps for solving the HSP. Several negative results on the shortcomings of this random strong method are known. In this paper however, we show that the random strong method can be quite powerful under certain conditions on the group G. We define a parameter r(G) for a group G and show that O((\log |G| / r(G))^2) iterations of the random strong method give enough classical information to identify a hidden subgroup in G. We illustrate the power of the random strong method via a concrete example of the HSP over finite Heisenberg groups. We show that r(G) = Ω(1) for these groups; hence the HSP can be solved using polynomially many random strong Fourier samplings followed by a possibly exponential classical post-processing without further queries. The quantum part of our algorithm consists of a polynomial computation followed by measuring in a random orthonormal basis. This gives the first example of a group where random representation bases do help in solving the HSP and for which no explicit representation bases are known that solve the problem with (\log G)^O(1) Fourier samplings. As an interesting by-product of our work, we get an algorithm for solving the state identification problem for a set of nearly orthogonal pure quantum states.

quant-ph↗
arXiv2005

Is partial quantum search of a database any easier?

In this paper, we consider the partial database search problem where given a database on N items, we are required to determine the first k bits of an address x such that f(x)=1. We derive an algorithm and a lower bound for this problem in the quantum circuits model. Let q(k,N) be the minimum number of queries needed to find the first k bits of the required address x. We show that there exist constants c_k and d_k such that (pi/4) (1 - d_k/sqrt{K}) sqrt{N} <= q(k,n) <= (pi/4) (1 - c_k/sqrt{K}) sqrt{N}, where K=2^k. Thus, it is always easier to determine a few bits of the target address than to find the entire address, but as k becomes large this advantage reduces rapidly.

quant-ph↗
arXiv2004

Quantum search for multiple items using parallel queries

In the quantum database search problem we are required to search for an item in a database. In this paper, we consider a generalization of this problem, where we are provided d identical copes of a database each with N items which we can query in parallel. Then, given k items, we are required to determine the locations where these items are stored. We show that any quantum algorithm for this task must perform Omega(sqrt{Nk/d min{d,k}}) parallel queries. We also design an algorithm whose performance comes within a factor O(log d) of this lower bound.

quant-ph↗
arXiv2003

A direct sum theorem in communication complexity via message compression

We prove lower bounds for the direct sum problem for two-party bounded error randomised multiple-round communication protocols. Our proofs use the notion of information cost of a protocol, as defined by Chakrabarti, Shi, Wirth and Yao and refined further by Bar-Yossef, Jayram, Kumar and Sivakumar. Our main technical result is a `compression' theorem saying that, for any probability distribution $μ$ over the inputs, a $k$-round private coin bounded error protocol for a function $f$ with information cost $c$ can be converted into a $k$-round deterministic protocol for $f$ with bounded distributional error and communication cost $O(kc)$. We prove this result using a substate theorem about relative entropy and a rejection sampling argument. Our direct sum result follows from this `compression' result via elementary information theoretic arguments. We also consider the direct sum problem in quantum communication. Using a probabilistic argument, we show that messages cannot be compressed in this manner even if they carry small information. Hence, new techniques may be necessary to tackle the direct sum problem in quantum communication.

cs.CC↗
arXiv2003

A lower bound for bounded round quantum communication complexity of set disjointness

We consider the class of functions whose value depends only on the intersection of the input X_1,X_2, ..., X_t; that is, for each F in this class there is an f_F: 2^{[n]} \to {0,1}, such that F(X_1,X_2, ..., X_t) = f_F(X_1 \cap X_2 \cap ... \cap X_t). We show that the t-party k-round communication complexity of F is Omega(s_m(f_F)/(k^2)), where s_m(f_F) stands for the `monotone sensitivity of f_F' and is defined by s_m(f_F) \defeq max_{S\subseteq [n]} |{i: f_F(S \cup {i}) \neq f_F(S)|. For two-party quantum communication protocols for the set disjointness problem, this implies that the two parties must exchange Omega(n/k^2) qubits. For k=1, our lower bound matches the Omega(n) lower bound observed by Buhrman and de Wolf (based on a result of Nayak, and for 2 <= k <= n^{1/4}, improves the lower bound of Omega(sqrt{n}) shown by Razborov. (For protocols with no restrictions on the number of rounds, we can conclude that the two parties must exchange Omega(n^{1/3}) qubits. This, however, falls short of the optimal Omega(sqrt{n}) lower bound shown by Razborov.)

quant-ph↗
arXiv2001

Depth-3 Arithmetic Circuits for S^2_n(X) and Extensions of the Graham-Pollack Theorem

We consider the problem of computing the second elementary symmetric polynomial S^2_n(X) using depth-three arithmetic circuits of the form "sum of products of linear forms". We consider this problem over several fields and determine EXACTLY the number of multiplication gates required. The lower bounds are proved for inhomogeneous circuits where the linear forms are allowed to have constants; the upper bounds are proved in the homogeneous model. For reals and rationals, the number of multiplication gates required is exactly n-1; in most other cases, it is \ceil{n/2}. This problem is related to the Graham-Pollack theorem in algebraic graph theory. In particular, our results answer the following question of Babai and Frankl: what is the minimum number of complete bipartite graphs required to cover each edge of a complete graph an odd number of times? We show that for infinitely many n, the answer is \ceil{n/2}.

cs.DM↗
arXiv2001

The Quantum Complexity of Set Membership

We study the quantum complexity of the static set membership problem: given a subset S (|S| \leq n) of a universe of size m (m \gg n), store it as a table of bits so that queries of the form `Is x \in S?' can be answered. The goal is to use a small table and yet answer queries using few bitprobes. This problem was considered recently by Buhrman, Miltersen, Radhakrishnan and Venkatesh, where lower and upper bounds were shown for this problem in the classical deterministic and randomized models. In this paper, we formulate this problem in the "quantum bitprobe model" and show tradeoff results between space and time.In this model, the storage scheme is classical but the query scheme is quantum.We show, roughly speaking, that similar lower bounds hold in the quantum model as in the classical model, which imply that the classical upper bounds are more or less tight even in the quantum case. Our lower bounds are proved using linear algebraic techniques.

quant-ph↗
arXiv2000

Analytical Studies of Strategies for Utilization of Cache Memory in Computers

We analyze quantitatively several strategies for better utilization of the {\em cache} or the {\em {fast access}} memory in computers. We define a performance factor $α$ that denotes the fraction of the cache area utilized when the main memory is accessed at random. We calculate $α$ exactly for different competing strategies, including the hash-rehash and the skewed-associative strategies which were earlier analyzed via simulations.

cond-mat.stat-mech↗