Search

Multiple groupcasting over the broadcast channel (BC) is studied. In particular, an inner bound is obtained for the $K$-receiver discrete memoryless (DM) BC for the diamond message set which consists of four groupcast messages: one desired by all receivers, one by all but two receivers, and two more desired by all but each one of those two receivers. The inner bound is based on rate-splitting and superposition coding and is given in explicit form herein as a union over coding distributions of four-dimensional polytopes. When specialized to the so-called combination network, which is a class of three-layer (two-hop) broadcast networks parameterized by $2^K-1$ finite-and-arbitrary-capacity noiseless links from the source node in the first layer to as many nodes of the second layer, our top-down approach from the DM BC to the combination network yields an explicit inner bound as a single polytope via the identification of a single coding distribution. This inner bound consists of inequalities which are then identified to be within the class of a plethora of (indeed, infinitely many) generalized cut-set outer bounds recently obtained by Salimi et al for broadcast networks. We hence establish the capacity region of the general $K$-user combination network for the diamond message set, and do so in explicit form. Such a result implies a certain strength of our inner bound for the DM BC in that it (a) produces a hitherto unknown capacity region when specialized to the combination network and (b) may capture many combinatorial aspects of the capacity region of the $K$-receiver DM BC itself (for the diamond message set). Moreover, we further extend that inner bound by adding binning to it and providing that inner bound also in explicit form as a union over coding distributions of four-dimensional polytopes in the message rates., Comment: 23 pages. Under review, IEEE Transactions on Information Theory. This paper was presented in part at the 2020 IEEE Intnl. Symp. Information Theory, Los Angeles, CA, 2020

A general inner bound is given for the discrete memoryless broadcast channel with an arbitrary number of users and general message sets, a setting that accounts for the most general form of concurrent groupcasting, with messages intended for any set of subsets of receivers. Achievability is based on superposition coding and rate-splitting without and with binning, where each receiver jointly decodes both its desired messages as well as the partial interference assigned to it via rate-splitting. The proof of achievability builds on the techniques for the description and analysis of superposition coding recently developed by the authors for the multiple-access channel with general messages as well as a new recursive mutual covering lemma for the analysis of the more general achievable scheme with binning., Comment: 8 pages. This work was presented in part at the 2017 IEEE International Symposium on Information Theory, Aachen, Germany, 2017, and appears in part in that symposium's proceedings. It was performed when the first author was a doctoral student at the University of Colorado, Boulder, and appears in his Ph.D. thesis

The K-user discrete memoryless (DM) broadcast channel (BC) with two nested multicast messages is studied in which one common message is to be multicast to all receivers and the second private message to a subset of receivers. The receivers that must decode both messages are referred to as private receivers and the others that must decode only the common message as common receivers. For two nested multicast messages, we establish the capacity region for several classes of DM BCs characterized by the respective associated sets of pair-wise relationships between and among the common and private receivers, each described by the well-known more capable or less noisy conditions. For three classes of DM BCs, the capacity region is simply achieved by superposition coding and the proofs of the converses rely on a recently found information inequality. The achievable rate region is then enhanced through the addition of a splitting of the private message into as many parts as there are common receivers and indirect decoding. A closed-form two-dimensional polyhedral description is obtained for it for a given coding distribution. Through a converse result that relies on the well-known Csiszar sum lemma and the information inequality, a specialization of this region that involves splitting the private message into just two sub-messages is proved to be the capacity region for several classes of DM BCs, beyond those for which superposition coding alone is capacity optimal, thereby underscoring the benefit of rate-splitting. All previously known capacity results for DM BCs with two nested multicast messages for the two and three-receiver DM BCs as well as DM BCs with one private or one common receiver are included in the general framework presented in this work., Comment: 24 pages, 6 figures. This manuscript is under review for possible publication in the IEEE Transactions on Information Theory. It was presented in part at 2017 IEEE International Symposium on Information Theory, Aachen, Germany

A novel class of achievable rate regions is obtained for the K-receiver broadcast channel with two groupcast messages. The associated achievability schemes are parameterized by an expansion of the message set which then determines how random coding techniques are employed, which include generalized versions of {\em up-set} message-splitting, the generation of possibly multiple auxiliary codebooks for certain compositions of split messages using superposition coding, partial interference decoding at all receivers, and joint unique and non-unique decoding. New capacity results are established for certain partially ordered classes of general broadcast channels for certain two non-nested messages. Moreover, when specialized to the combination network (CN), some of the inner bounds are shown, via converse results, to result in the capacity region for (a) the two messages intended for two sets of K-1 receivers each and (b) two nested messages in which one message is intended for one or (c) two (common) receivers. In the latter two cases, we hence recover previous results by Bidokhti et al obtained therein using network coding schemes based on rate-splitting and linear superposition coding. Furthermore, we show the achievability of rate pairs in two examples of CNs, with three and four common receivers each, used in the previous literature to show the sub-optimality of rate-splitting and linear superposition coding, and to motivate a pre-encoding technique and a block-Markov linear superposition coding for the CN, with the latter then lifted to the general broadcast channel. Our results suggest that the proposed framework here, when specialized to the CN, is strong enough to incorporate the enhancements afforded by those two latter techniques, thereby suggesting among other things, that perhaps block-Markov superposition coding is not necessary for the general broadcast channel., Comment: 45 pages, 5 figures. The overall perspective and the novelty of results are clarified and several new results have been added relative to the original submission. This manuscript is under second round of review for possible publication in the IEEE Transactions on Information Theory. It was presented in part at the 2018 IEEE International Symposium on Information Theory in Vail, Colorado

The generalized degrees of freedom (GDoF) region of the multiple-input multiple-output (MIMO) Gaussian Z-interference channel with an arbitrary number of antennas at each node is established under the assumption of delayed channel state information at transmitters (CSIT). The GDoF region is parameterized by $\alpha$, which links the interference-to-noise ratio (INR) to the signal-to-noise ratio (SNR) via $INR=SNR^{\alpha}$. A new outer bound for the GDoF region is established by maximizing a bound on the weighted sum-rate of the two users, which in turn is obtained by using a combination of genie-aided side-information and an extremal inequality. The maximum weighted sum-rate in the high SNR regime is shown to occur when the transmission covariance matrix of the interfering transmitter has full rank. An achievability scheme based on block-Markov encoding and backward decoding is developed which uses interference quantization and digital multicasting to take advantage of the channel statistics of the cross-link, and the scheme is separately shown to be GDoF-optimal in both the weak ($\alpha\leq1$) and strong ($\alpha>1$) interference regimes. This is the first complete characterization of the GDoF region of any interference network with delayed CSIT, as well as the first such GDoF characterization of a MIMO network with delayed CSIT and arbitrary number of antennas at each node. For all antenna tuples, the GDoF region is shown to be equal to or larger than the degrees of freedom (DoF) region over the entire range of $\alpha$, which leads to a V-shaped maximum sum-GDoF as a function of $\alpha$, with the minimum occurring at $\alpha=1$. The delayed CSIT GDoF region and the sum-DoF are compared with their counterparts under perfect CSIT, thereby characterizing all antenna tuples and ranges of $\alpha$ for which delayed CSIT is sufficient to achieve the perfect CSIT GDoF region or sum-DoF., Comment: submitted, IEEE Transactions on Information Theory

Approximate capacity regions are established for a class of interfering multiple access channels consisting of two multiple-access channels (MACs), each with an arbitrary number of transmitters, with one transmitter in each MAC causing interference to the receiver of the other MAC, a channel we refer to henceforth as the MAC-IC-MAC. For the discrete memoryless (DM) MAC-IC-MAC, two inner bounds are obtained that are generalizations of prior inner bounds for the two-user DM interference channel (IC) due to Chong {\em et al}. For the semi-deterministic MAC-IC-MAC, it is shown that single-user coding at the non-interfering transmitters and superposition coding at the interfering transmitter of each MAC achieves a rate region that is within a quantifiable gap of the capacity region, thereby extending such a result for the two-user semi-deterministic IC by Telatar and Tse. For the Gaussian MAC-IC-MAC, an approximate capacity region that is within a constant gap of the capacity region is obtained, generalizing such a result for the two-user Gaussian IC by Etkin {\em et al}. Contrary to the aforementioned approximate capacity results for the two-user IC whose achievability requires the union of all admissible input distributions, our gap results on the semi-deterministic and the Gaussian MAC-IC-MAC are achievable by only a subset and one of all admissible coding distributions, respectively. The symmetric generalized degrees of freedom (GDoF) of the symmetric Gaussian MAC-IC-MAC with more than one user per cell, which is a function of the interference strength (the ratio of INR to SNR at high SNR, both expressed in dB) and the numbers of users in each cell, is V-shaped with flat shoulders. ...

We study the degrees of freedom (DoF) regions of the two-user multiple-input multiple-output (MIMO) broadcast channel with a general message set (BC-CM) - that includes private and common messages - under fast fading. Nine different channel state knowledge assumptions -collectively known as hybrid CSIT models - are considered wherein the transmitter has either perfect/instantaneous (P), delayed (D) or no (N) channel state information (CSI) from each of the two receivers. General antenna configurations are addressed wherein the three terminals have arbitrary numbers of antennas. The DoF regions are established for the five hybrid CSIT models in which either both channels are unknown at the transmitter or each of the two channels is known perfectly or with delay. In the four remaining cases in which exactly one of the two channels is unknown at the transmitter, the DoF regions under the restriction of linear encoding strategies - also known as the linear DoF (LDoF) regions- are established. As the key to the converse proofs of the LDoF region of the MIMO BC-CM under such hybrid CSIT assumptions, we show that, when only considering linear encoding strategies, the channel state information from the receiver with more antennas does not help if there is no channel state information available from the receiver with fewer antennas. This result is conjectured to be true even without the restriction on the encoding strategies to be linear. If true, the LDoF regions obtained for the four hybrid CSIT cases herein will also be the DoF regions for those cases., Comment: submitted to T-IT

We establish the DoF region for the MIMO 2x2 interference network with a general message set, consisting of nine messages, one for each pair of a subset of transmitters at which that message is known and a subset of receivers where that message is desired. An outer bound on the general nine-message network is obtained and then it is shown to be tight, establishing the DoF region for the most general antenna setting wherein all four nodes have an arbitrary number of antennas each. The DoF-optimal scheme is applicable to the MIMO 2x2 network with constant channel coefficients, and hence, a fortiori, to time/frequency varying channel scenarios. In particular, a linear precoding scheme is proposed that can achieve all the DoF tuples in the DoF region. In it, the precise roles played by transmit zero-forcing, interference alignment, random beamforming, symbol extensions and asymmetric complex signaling are delineated. For instance, we identify a class of antenna settings in which ACS is required to achieve the fractional-valued corner points. Evidently, the DoF regions of all previously unknown cases of the 2x2 interference network with a subset of the nine-messages are established as special cases of the general result of this paper. In particular, the DoF region of the well-known four-message (and even three-message) MIMO X channel is established. This problem had remained open despite previous studies which had found inner and outer bounds that were not tight in general. Hence, the DoF regions of all special cases obtained from the general DoF region of the nine-message 2x2 interference network of this work that include at least three of the four X channel messages are new, among many others. Our work sheds light on how the same physical 2x2 network could be used by a suitable choice of message sets to take most advantage of the channel resource in a flexible and efficient manner., Comment: submitted to T-IT on Mar 4th, 2016

An outer bound for the degrees of freedom (DoF) region of the K-user multiple-input single-output (MISO) broadcast channel (BC) is developed under the hybrid channel state information at transmitter (CSIT) model, in which the transmitter has instantaneous CSIT of channels to a subset of the receivers and delayed CSIT of channels to the rest of the receivers. For the 3-user MISO BC, when the transmitter has instantaneous CSIT of the channel to one receiver and delayed CSIT of channels to the other two, two new communication schemes are designed, which are able to achieve the DoF tuple of $\left(1,\frac{1}{3},\frac{1}{3}\right)$, with a sum DoF of $\frac{5}{3}$, that is greater than the sum DoF achievable only with delayed CSIT. Another communication scheme showing the benefit of the alternating CSIT model is also developed, to obtain the DoF tuple of $\left(1,\frac{4}{9},\frac{4}{9}\right)$ for the 3-user MISO BC.

We introduce the notion of Universally Decodable Matrices of Genus g (UDMG), which for g=0 reduces to the notion of Universally Decodable Matrices (UDM) introduced in [8]. A UDMG is a set of L matrices over a finite field, each with K rows, and a linear independence condition satisfied by collections of K+g columns formed from the initial segments of the matrices. We consider the mathematical structure of UDMGs and their relation to linear vector codes. We then give a construction of UDMG based on curves of genus g over the finite field, which is a natural generalization of the UDM constructed in [8]. We provide upper (and constructable lower) bounds for L in terms of K, q, g, and the number of columns of the matrices. We will show there is a fundamental trade off (Theorem 5.4) between L and g, akin to the Singleton bound for the minimal Hamming distance of linear vector codes., Comment: 23 pages

The question of whether the degrees of freedom (DoF) of multi-user networks can be enhanced even under isotropic fading and no channel state information (or output feedback) at the transmitters (CSIT) is investigated. Toward this end, the two-user MIMO (multiple-input, multiple-output) broadcast and interference channels are studied with no side-information whatsoever at the transmitters and with receivers equipped with full-duplex radios. The full-duplex feature allows for receiver cooperation because each receiver, in addition to receiving the signals sent by the transmitters, can also simultaneously transmit a signal in the same band to the other receiver. Unlike the case of MIMO networks with CSIT and full-duplex receivers, for which DoF are known, it is shown that for MIMO networks with no CSIT, full-duplex receiver cooperation is beneficial to such an extent that even the DoF region is enhanced. Indeed, for important classes of two-user MIMO broadcast and interference channels, defined by certain relationships on numbers of antennas at different terminals, the exact DoF regions are established. The key to achieving DoF-optimal performance for such networks are new retro-cooperative interference alignment schemes. Their optimality is established via the DoF analysis of certain genie-aided or enhanced version of those networks., Comment: This work was presented at the Workshop on Interference in Wireless Networks, Boston University, June 2012

The degrees of freedom (DoF) region of the two-user MIMO (multiple-input multiple-output) interference channel is established under a new model termed as hybrid CSIT. In this model, one transmitter has delayed channel state information (CSI) and the other transmitter has instantaneous CSIT, of incoming channel matrices at the respective unpaired receivers, and neither transmitter has any knowledge of the incoming channel matrices of its respective paired receiver. The DoF region for hybrid CSIT, and consequently that of $2\times2\times3^{5}$ CSIT models, is completely characterized, and a new achievable scheme based on a combination of transmit beamforming and retrospective interference alignment is developed. Conditions are obtained on the numbers of antennas at each of the four terminals such that the DoF region under hybrid CSIT is equal to that under (a) global and instantaneous CSIT and (b) global and delayed CSIT, with the remaining cases resulting in a DoF region with hybrid CSIT that lies somewhere in between the DoF regions under the instantaneous and delayed CSIT settings. Further synergistic benefits accruing from switching between the two hybrid CSIT models are also explored.

The interference channel with common information (IC-CI) consists of two transmit-receive pairs that communicate over a common noisy medium. Each transmitter has an individual message for its paired receiver, and additionally, both transmitters have a common message to deliver to both receivers. In this paper, through explicit inner and outer bounds on the capacity region, we establish the capacity region of the Gaussian IC-CI to within a bounded gap of one bit, independently of the values of all channel parameters. Using this constant-gap characterization, the generalized degrees of freedom (GDoF) region is determined. It is shown that the introduction of the common message leads to an increase in the GDoF over that achievable over the Gaussian interference channel without a common message, and hence to an unbounded improvement in the achievable rate. A surprising feature of the capacity-within-one-bit result is that most of the available benefit (i.e., to within one bit of capacity) due to the common message is achieved through a simple and explicit coding scheme that involves independent signaling at the two transmitters so that, in effect, this scheme forgoes the opportunity for transmitter cooperation that is inherently available due to shared knowledge of the common message at both transmitters., Comment: Submitted to IEEE Trans. on Information Theory, Feb. 2012

The two-user multiple-input multiple-output (MIMO) fast-fading interference channel (IC) with an arbitrary number of antennas at each of the four terminals is studied under the settings of Shannon feedback, limited Shannon feedback, and output feedback, wherein all or certain channel matrices and outputs, or just the channel outputs, respectively, are available to the transmitters with a finite delay. While for most numbers of antennas at the four terminals, it is shown that the DoF regions with Shannon feedback and for the limited Shannon feedback settings considered here are identical, and equal to the DoF region with just delayed channel state information (CSIT), it is shown that this is not always the case. For a specific class of MIMO ICs characterized by a certain relationship between the numbers of antennas at the four nodes, the DoF regions with Shannon and the limited Shannon feedback settings, while again being identical, are strictly bigger than the DoF region with just delayed CSIT. To realize these DoF gains with Shannon or limited Shannon feedback, a new retrospective interference alignment scheme is developed wherein transmitter cooperation made possible by output feedback in addition to delayed CSIT is employed to effect a more efficient form of interference alignment than is feasible with previously known schemes that use just delayed CSIT. The DoF region for just output feedback, in which each transmitter has delayed knowledge of only the receivers' outputs, is also obtained for all but a class of MIMO ICs that satisfy one of two inequalities involving the numbers of antennas., Comment: 30 pages, 3 tables, 9 figures. This paper was submitted to the IEEE Trans. Inform. Th. Oct. 2011. It was presented in part at the 49th Annual Allerton Conference on Communications, Control and Computing in Sept. 2011

The layered two-hop, two-flow interference network is considered that consists of two sources, two relays and two destinations with the first hop network between he sources and the relays and the second hop network between relays and destinations both being i.i.d. Rayleigh fading Gaussian interference channels. Two feedback models are studied. In the first one, called the delayed channel state information at the sources (delayed CSI-S) model, the sources know all channel coefficients with a finite delay but the relays have no side information whatsoever. In the second feedback model, referred to as the limited Shannon feedback model, the relays know first hop channel coefficients instantaneously and the second hop channel with a finite delay and one relay knows the received signal of one of the destinations with a finite delay and the other relay knows the received signal of the other destination with a finite delay but there is no side information at the sources whatsoever. It is shown in this paper that under both these settings, the layered two-hop, two-flow interference channel has 4/3 degrees of freedom. The result is obtained by developing a broadcast-channel-type upper-bound and new achievability schemes based on the ideas of retrospective interference alignment and retro-cooperative interference alignment, respectively., Comment: Submitted, July 10, 2011 to the 2011 Allerton Conf. Commun., Control, Comput., Monticello, IL; accepted August 04, 2011

The fundamental diversity-multiplexing tradeoff of the three-node, multi-input, multi-output (MIMO), quasi-static, Rayleigh faded, half-duplex relay channel is characterized for an arbitrary number of antennas at each node and in which opportunistic scheduling (or dynamic operation) of the relay is allowed, i.e., the relay can switch between receive and transmit modes at a channel dependent time. In this most general case, the diversity-multiplexing tradeoff is characterized as a solution to a simple, two-variable optimization problem. This problem is then solved in closed form for special classes of channels defined by certain restrictions on the numbers of antennas at the three nodes. The key mathematical tool developed here that enables the explicit characterization of the diversity-multiplexing tradeoff is the joint eigenvalue distribution of three mutually correlated random Wishart matrices. Previously, without actually characterizing the diversity-multiplexing tradeoff, the optimality in this tradeoff metric of the dynamic compress-and-forward (DCF) protocol based on the classical compress-and-forward scheme of Cover and El Gamal was shown by Yuksel and Erkip. However, this scheme requires global channel state information (CSI) at the relay. In this work, the so-called quantize-map and forward (QMF) coding scheme due to Avestimehr {\em et} {\em al} is adopted as the achievability scheme with the added benefit that it achieves optimal tradeoff with only the knowledge of the (channel dependent) switching time at the relay node. Moreover, in special classes of the MIMO half-duplex relay channel, the optimal tradeoff is shown to be attainable even without this knowledge. Such a result was previously known only for the half-duplex relay channel with a single antenna at each node, also via the QMF scheme., Comment: 42 pages, 7 figures; submitted to the Trans. of IT

The two-user multi-input, multi-output (MIMO) interference and cognitive radio channels are studied under the assumption of no channel state information at the transmitter (CSIT) from the degrees of freedom (DoF) region perspective. With $M_i$ and $N_i$ denoting the number of antennas at transmitter $i$ and receiver $i$ respectively, the DoF regions of the MIMO interference channel were recently characterized by Huang et al., Zhu and Guo, and by the authors of this paper for all values of numbers of antennas except when $\min(M_1,N_1) > N_2 > M_2$ (or $\min(M_2,N_2) > N_1 > M_1$). This latter case was solved more recently by Zhu and Guo who provided a tight outer-bound. Here, a simpler and more widely applicable proof of that outer-bound is given based on the idea of interference localization. Using it, the DoF region is also established for the class of MIMO cognitive radio channels when $\min(M_1+M_2,N_1) > N_2 > M_2$ (with the second transmitter cognitive) -- the only class for which the inner and outer bounds previously obtained by the authors were not tight -- thereby completing the DoF region characterization of the general 2-user MIMO cognitive radio channel as well., Comment: Submitted to IEEE Trans. Information Theory, May 2011. A material in this paper will be presented in part at the IEEE Intern. Symp. Information Theory (ISIT), Aug. 2011

The generalized degrees of freedom (GDoF) region of the MIMO Gaussian interference channel (IC) is obtained for the general case of an arbitrary number of antennas at each node and where the signal-to-noise ratios (SNR) and interference-to-noise ratios (INR) vary with arbitrary exponents to a nominal SNR. The GDoF region reveals various insights through the joint dependence of optimal interference management techniques (at high SNR) on the SNR exponents that determine the relative strengths of direct-link SNRs and cross-link INRs and the numbers of antennas at the four terminals. For instance, it permits an in-depth look at the issue of rate-splitting and partial decoding and it reveals that, unlike in the scalar IC, treating interference as noise is not always GDoF-optimal even in the very weak interference regime. Moreover, while the DoF-optimal strategy that relies just on transmit/receive zero-forcing beamforming and time-sharing is not GDoF optimal (and thus has an unbounded gap to capacity), the precise characterization of the very strong interference regime -- where single-user DoF performance can be achieved simultaneously for both users-- depends on the relative numbers of antennas at the four terminals and thus deviates from what it is in the SISO case. For asymmetric numbers of antennas at the four nodes the shape of the symmetric GDoF curve can be a "distorted W" curve to the extent that for certain MIMO ICs it is a "V" curve., Comment: 38 pages, 14 figures. Submitted to Trans. of IT

The generalized degrees of freedom (GDoF) region of the MIMO Gaussian interference channel is obtained for the general case with an arbitrary number of antennas at each node and where the SNR and interference-to-noise ratios (INRs) vary with arbitrary exponents to a nominal SNR. The GDoF region reveals various insights through the joint dependence of optimal interference management techniques at high SNR on the SNR exponents that determine the relative strengths of direct-link SNRs and cross-link INRs and the numbers of antennas at the four terminals. For instance, it permits an in-depth look at the issue of rate-splitting and partial decoding at high SNR and it reveals that, unlike in the SISO case, treating interference as noise is not GDoF optimal always even in the very weak interference regime. Moreover, while the DoF-optimal strategy that relies just on transmit/receive zero-forcing beamforming and time-sharing is not GDoF optimal (and thus has an unbounded gap to capacity) the precise characterization of the very strong interference regime, where single-user DoF performance can be achieved simultaneously for both users, depends on the relative numbers of antennas at the four terminals and thus deviates from what it is in the SISO case. For asymmetric numbers of antennas at the four nodes the shape of the symmetric GDoF curve can be a "distorted W" curve to the extent that for certain MIMO ICs it is a "V" curve., Comment: 5 pages, 8 figures; submitted to ISIT 2011

The capacity region of the 2-user multi-input multi-output (MIMO) Gaussian interference channel (IC) is characterized to within a constant gap that is independent of the channel matrices for the general case of the MIMO IC with an arbitrary number of antennas at each node. An achievable rate region and an outer bound to the capacity region of a class of interference channels were obtained in previous work by Telatar and Tse as unions over all possible input distributions. In contrast to that previous work on the MIMO IC, a simple and an explicit achievable coding scheme are obtained here and shown to have the constant-gap-to-capacity property and in which the sub-rates of the common and private messages of each user are explicitly specified for each achievable rate pair. The constant-gap-to-capacity results are thus proved in this work by first establishing explicit upper and lower bounds to the capacity region. A reciprocity result is also proved which is that the capacity of the reciprocal MIMO IC is within a constant gap of the capacity region of the forward MIMO IC., Comment: 22 pages, 5 figures, accepted in Trans. on Inform. Th

The degrees of freedom (DoF) region of the 2-user multiple-antenna or MIMO (multiple-input, multiple-output) interference channel (IC) is studied under fast fading and the assumption of {\em delayed} channel state information (CSI) wherein all terminals know all (or certain) channel matrices perfectly, but with a delay, and each receiver in addition knows its own incoming channels instantaneously. The general MIMO IC is considered with an arbitrary number of antennas at each of the four terminals. Dividing it into several classes depending on the relation between the numbers of antennas at the four terminals, the fundamental DoF regions are characterized under the delayed CSI assumption for {\em all} possible values of number of antennas at the four terminals. In particular, an outer bound on the DoF region of the general MIMO IC is derived. This bound is then shown to be tight for all MIMO ICs by developing interference alignment based achievability schemes for each class. A comparison of these DoF regions under the delayed CSI assumption is made with those of the idealistic `perfect CSI' assumption where perfect and instantaneous CSI is available at all terminals on the one hand and with the DoF regions of the conservative `no CSI' assumption on the other, where CSI is available at the receivers but not at all at the transmitters., Comment: New results are added. 57 pages, 6 figures, 2 tables, submitted to IEEE Trans. Inform. Theory

The degrees of freedom (DoF) region of the fast-fading MIMO (multiple-input multiple-output) Gaussian broadcast channel (BC) is studied when there is delayed channel state information at the transmitter (CSIT). In this setting, the channel matrices are assumed to vary independently across time and the transmitter is assumed to know the channel matrices with some arbitrary finite delay. An outer-bound to the DoF region of the general $K$-user MIMO BC (with an arbitrary number of antennas at each terminal) is derived. This outer-bound is then shown to be tight for two classes of MIMO BCs, namely, (a) the two-user MIMO BC with arbitrary number of antennas at all terminals, and (b) for certain three-user MIMO BCs where all three receivers have an equal number of antennas and the transmitter has no more than twice the number of antennas present at each receivers. The achievability results are obtained by developing an interference alignment scheme that optimally accounts for multiple, and possibly distinct, number of antennas at the receivers., Comment: 27 pages, 6 figures; submitted to IEEE Trans. on Information Theory, Dec. 2011

The diversity-multiplexing tradeoff of the dynamic decode-and-forward protocol is characterized for the half-duplex three-terminal (m,k,n)-relay channel where the source, relay and the destination terminals have m, k and n antennas, respectively. It is obtained as a solution to a simple, two-variable, convex optimization problem and this problem is solved in closed form for special classes of relay channels, namely, the (1,k,1) relay channel, the (n,1,n) relay channel and the (2,k,2) relay channel. Moreover, the tradeoff curves for a certain class of relay channels, such as the (m,k,n>k) channels, are identical to those for the decode-and-forward protocol for the full duplex channel while for other classes of channels they are marginally lower at high multiplexing gains. Our results also show that for some classes of relay channels and at low multiplexing gains the diversity orders of the dynamic decode-and-forward protocol protocol are greater than those of the static compress-and-forward protocol which in turn is known to be tradeoff optimal over all {\em static} half duplex protocols. In general, the dynamic decode-and-forward protocol has a performance that is comparable to that of the static compress-and-forward protocol which, unlike the dynamic decode-and-forward protocol, requires global channel state information at the relay node. Its performance is also close to that of the decode-and-forward protocol over the full-duplex relay channel thereby indicating that the half-duplex constraint can be compensated for by the dynamic operation of the relay wherein the relay switches from the receive to the transmit mode based on the source-relay channel quality., Comment: 37 pages and 7 figures

The fundamental diversity-multiplexing tradeoff (DMT) of the quasi-static fading, symmetric $2$-user MIMO interference channel (IC) with channel state information at the transmitters (CSIT) and a short term average power constraint is obtained. The general case is considered where the interference-to-noise ratio (INR) at each receiver scales differently from the signal-to-noise ratio (SNR) at the receivers. The achievability of the DMT is proved by showing that a simple Han-Kobayashi coding scheme can achieve a rate region which is within a constant (independent of SNR) number of bits from a set of upper bounds to the capacity region of the IC. In general, only part of the DMT curve with CSIT can be achieved by coding schemes which do not use any CSIT (No-CSIT). A result in this paper establishes a threshold for the INR beyond which the DMT with CSIT coincides with that with No-CSIT. Our result also settles one of the conjectures made in~\cite{EaOlCv}. Furthermore, the fundamental DMT of a class of non-symmetric ICs with No-CSIT is also obtained wherein the two receivers have different numbers of antennas., Comment: 5 pages, 3 figures. To be presented at ISIT 2010

The fundamental generalized diversity-multiplexing tradeoff (GDMT) of the quasi-static fading MIMO Z interference channel (Z-IC) is established for the general Z-IC with an arbitrary number of antennas at each node under the assumptions of full channel state information at the transmitters (CSIT) and a short-term average power constraint. In the GDMT framework, the direct link signal-to-noise ratios (SNR) and cross-link interference-to-noise ratio (INR) are allowed to vary so that their ratios relative to a nominal SNR in the dB scale, i.e., the SNR/INR exponents, are fixed. It is shown that a simple Han-Kobayashi message-splitting/partial interference decoding scheme that uses only partial CSIT -- in which the second transmitter's signal depends only on its cross-link channel matrix and the first user's transmit signal doesn't need any CSIT whatsoever -- can achieve the full-CSIT GDMT of the MIMO Z-IC. The GDMT of the MIMO Z-IC under the No-CSIT assumption is also obtained for some range of multiplexing gains. The size of this range depends on the numbers of antennas at the four nodes and the SNR and INR exponents of the direct and cross links, respectively. For certain classes of channels including those in which the interfered receiver has more antennas than do the other nodes, or when the INR exponent is greater than a certain threshold, the GDMT of the MIMO Z-IC under the No-CSIT assumption is completely characterized., Comment: Submitted to the Transactions of Information Theory, 34 pages, 6 figures

This paper has been withdrawn due to an error in one of the equations in the extended portion., Comment: This paper has been withdrawn.

The Gaussian broadcast channel (GBC) with $K$ transmit antennas and $K$ single-antenna users is considered for the case in which the channel state information is obtained at the transmitter via a finite-rate feedback link of capacity $r$ bits per user. The throughput (i.e., the sum-rate normalized by $K$) of the GBC is analyzed in the limit as $K \to \infty$ with $\frac{r}{K} \to \bar{r}$. Considering the transmission strategy of zeroforcing dirty paper coding (ZFDPC), a closed-form expression for the asymptotic throughput is derived. It is observed that, even under the finite-rate feedback setting, ZFDPC achieves a significantly higher throughput than zeroforcing beamforming. Using the asymptotic throughput expression, the problem of obtaining the number of users to be selected in order to maximize the throughput is solved., Comment: Submitted to ISIT 2010

This paper has been withdrawn by the author(s) for revision., Comment: This paper has been withdrawn

This paper has been withdrawn by the author(s) for revision., Comment: This paper has been withdrawn

While interference alignment schemes have been employed to realize the full multiplexing gain of $K$-user interference channels, the analyses performed so far have predominantly focused on the case when global channel knowledge is available at each node of the network. This paper considers the problem where each receiver knows its channels from all the transmitters and feeds back this information using a limited number of bits to all other terminals. In particular, channel quantization over the composite Grassmann manifold is proposed and analyzed. It is shown, for $K$-user multiple-input, multiple-output (MIMO) interference channels, that when the transmitters use an interference alignment strategy as if the quantized channel estimates obtained via this limited feedback are perfect, the full sum degrees of freedom of the interference channel can be achieved as long as the feedback bit rate scales sufficiently fast with the signal-to-noise ratio. Moreover, this is only one extreme point of a continuous tradeoff between achievable degrees of freedom region and user feedback rate scalings which are allowed to be non-identical. It is seen that a slower scaling of feedback rate for any one user leads to commensurately fewer degrees of freedom for that user alone.

The degrees of freedom (DoF) regions are characterized for the multiple-input multiple-output (MIMO) broadcast channel (BC), interference channels (IC) (including X and multi-hop interference channels) and the cognitive radio channel (CRC), when there is perfect and no channel state information at the receivers and the transmitter(s) (CSIR and CSIT), respectively. For the K-user MIMO BC, the exact characterization of the DoF region is obtained, which shows that a simple time-division-based transmission scheme is DoF-region optimal. Using the techniques developed for the MIMO BC, the corresponding problems for the two-user MIMO IC and the CRC are addressed. For both of these channels, inner and outer bounds to the DoF region are obtained and are seen to coincide for a vast majority of the relative numbers of antennas at the four terminals, thereby characterizing DoF regions for all but a few cases. Finally, the DoF regions of the $K$-user MIMO IC, the CRC, and X networks are derived for certain classes of these networks, including the one where all transmitters have an equal number of antennas and so do all receivers. The results of this paper are derived for distributions of fading channel matrices and additive noises that are more general than those considered in other simultaneous related works. The DoF regions with and without CSIT are compared and conditions on the relative numbers of antennas at the terminals under which a lack of CSIT does, or does not, result in the loss of DoF are identified, thereby providing, on the one hand, simple and robust communication schemes that don't require CSIT but have the same DoF performance as their previously found CSIT counterparts, and on the other hand, identifying situations where CSI feedback to transmitters would provide gains that are significant enough that even the DoF performance could be improved., Comment: 49 pages, 11 figures, under review, IEEE Trans. Inform. Th. Submitted Sept. 2009, Revised Jan. 2011

A Dirty Paper Coding (DPC) based transmission scheme for the Gaussian multiple-input multiple-output (MIMO) cognitive radio channel (CRC) is studied when there is imperfect and perfect channel knowledge at the transmitters (CSIT) and the receivers, respectively. In particular, the problem of optimizing the sum-rate of the MIMO CRC over the transmit covariance matrices is dealt with. Such an optimization, under the DPC-based transmission strategy, needs to be performed jointly with an optimization over the inflation factor. To this end, first the problem of determination of inflation factor over the MIMO channel $Y=H_1 X + H_2 S + Z$ with imperfect CSIT is investigated. For this problem, two iterative algorithms, which generalize the corresponding algorithms proposed for the channel $Y=H(X+S)+Z$, are developed. Later, the necessary conditions for maximizing the sum-rate of the MIMO CRC over the transmit covariances for a given choice of inflation factor are derived. Using these necessary conditions and the algorithms for the determination of the inflation factor, an iterative, numerical algorithm for the joint optimization is proposed. Some interesting observations are made from the numerical results obtained from the algorithm. Furthermore, the high-SNR sum-rate scaling factor achievable over the CRC with imperfect CSIT is obtained., Comment: To be presented at ISIT 2009, Seoul, S. Korea

The problem of dirty paper coding (DPC) over the (multi-antenna) fading dirty paper channel (FDPC) Y = H(X + S) + Z is considered when there is imperfect knowledge of the channel state information H at the transmitter (CSIT). The case of FDPC with positive definite (p.d.) input covariance matrix was studied by the authors in a recent paper, and here the more general case of positive semi-definite (p.s.d.) input covariance is dealt with. Towards this end, the choice of auxiliary random variable is modified. The algorithms for determination of inflation factor proposed in the p.d. case are then generalized to the case of p.s.d. input covariance. Subsequently, the largest DPC-achievable high-SNR (signal-to-noise ratio) scaling factor over the no-CSIT FDPC with p.s.d. input covariance matrix is derived. This scaling factor is seen to be a non-trivial generalization of the one achieved for the p.d. case. Next, in the limit of low SNR, it is proved that the choice of all-zero inflation factor (thus treating interference as noise) is optimal in the 'ratio' sense, regardless of the covariance matrix used. Further, in the p.d. covariance case, the inflation factor optimal at high SNR is obtained when the number of transmit antennas is greater than the number of receive antennas, with the other case having been already considered in the earlier paper. Finally, the problem of joint optimization of the input covariance matrix and the inflation factor is dealt with, and an iterative numerical algorithm is developed., Comment: Presented at the 43rd Annual Conference on Information Sciences and Systems, John Hopkins University, March 2009

The problem of Dirty Paper Coding (DPC) over the Fading Dirty Paper Channel (FDPC) Y = H(X + S)+Z, a more general version of Costa's channel, is studied for the case in which there is partial and perfect knowledge of the fading process H at the transmitter (CSIT) and the receiver (CSIR), respectively. A key step in this problem is to determine the optimal inflation factor (under Costa's choice of auxiliary random variable) when there is only partial CSIT. Towards this end, two iterative numerical algorithms are proposed. Both of these algorithms are seen to yield a good choice for the inflation factor. Finally, the high-SNR (signal-to-noise ratio) behavior of the achievable rate over the FDPC is dealt with. It is proved that FDPC (with t transmit and r receive antennas) achieves the largest possible scaling factor of min(t,r) log SNR even with no CSIT. Furthermore, in the high SNR regime, the optimality of Costa's choice of auxiliary random variable is established even when there is partial (or no) CSIT in the special case of FDPC with t <= r. Using the high-SNR scaling-law result of the FDPC (mentioned before), it is shown that a DPC-based multi-user transmission strategy, unlike other beamforming-based multi-user strategies, can achieve a single-user sum-rate scaling factor over the multiple-input multiple-output Gaussian Broadcast Channel with partial (or no) CSIT., Comment: 5 pages with 2 figures, presented at 42nd Asilomar Conference on Signals, Systems, and Computers, Pacific Grove, USA, Oct. 2008

Reliable communication over the discrete-input/continuous-output noncoherent multiple-input multiple-output (MIMO) Rayleigh block fading channel is considered when the signal-to-noise ratio (SNR) per degree of freedom is low. Two key problems are posed and solved to obtain the optimum discrete input. In both problems, the average and peak power per space-time slot of the input constellation are constrained. In the first one, the peak power to average power ratio (PPAPR) of the input constellation is held fixed, while in the second problem, the peak power is fixed independently of the average power. In the first PPAPR-constrained problem, the mutual information, which grows as O(SNR^2), is maximized up to second order in SNR. In the second peak-constrained problem, where the mutual information behaves as O(SNR), the structure of constellations that are optimal up to first order, or equivalently, that minimize energy/bit, are explicitly characterized. Furthermore, among constellations that are first-order optimal, those that maximize the mutual information up to second order, or equivalently, the wideband slope, are characterized. In both PPAPR-constrained and peak-constrained problems, the optimal constellations are obtained in closed-form as solutions to non-convex optimizations, and interestingly, they are found to be identical. Due to its special structure, the common solution is referred to as Space Time Orthogonal Rank one Modulation, or STORM. In both problems, it is seen that STORM provides a sharp characterization of the behavior of noncoherent MIMO capacity., Comment: 52 pages, to appear in IEEE Transactions on Information Theory