\n \n \n
\n
\n\n \n \n J. Sebastian; and S. N. Diggavi.\n\n\n \n \n \n \n Generalized Degrees Freedom of Noncoherent MIMO Channels With Asymmetric Link Strengths.\n \n \n \n\n\n \n\n\n\n
IEEE Transactions on Information Theory, 66(7): 4431-4448. July 2020.\n
\n\n
\n\n
\n\n
\n\n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{9023963,\n abstract = {We study the generalized degrees of freedom (gDoF) of block-fading noncoherent multiple input multiple output (MIMO) channels with asymmetric distributions of link strengths and a coherence time of T symbol durations. We derive the optimal signaling structure for communication for the asymmetric MIMO channel, which is distinct from that for the MIMO channel with independent and identically distributed (i.i.d.) links. We extend the existing results for the single input multiple output (SIMO) channel with i.i.d. links to the asymmetric case, proving that selecting the statistically best antenna is gDoFoptimal. Using the gDoF result for the SIMO channel, we prove that for T = 1, the gDoF is zero for MIMO channels with arbitrary link strengths. We show that selecting the statistically best antenna is gDoF-optimal for the multiple input single output (MISO) channel. We also derive the gDoF for the 2 x 2 MIMO channel with different exponents in the direct and cross links. In this setting, we show that it is always necessary to use both the antennas to achieve the gDoF, in contrast to the results for the 2 x 2 MIMO channel with i.i.d. links. We show that having weaker crosslinks, gives gDoF gain compared to the case with i.i.d. links. For the noncoherent MIMO channel with i.i.d. links, the traditional method of training each transmit antenna independently is degrees of freedom (DoF) optimal, whereas we observe that for the asymmetric 2 x 2 MIMO channel, the traditional training is not gDoF-optimal. We extend this observation to a larger MxM MIMO channel by demonstrating a strategy that can achieve larger gDoF than a traditional trainingbased method.},\n author = {J. {Sebastian} and S. N. {Diggavi}},\n doi = {10.1109/TIT.2020.2978183},\n issn = {1557-9654},\n journal = {IEEE Transactions on Information Theory},\n keywords = {MIMO communication;MISO communication;SIMO communication;transmitting antennas;generalized degrees freedom;noncoherent MIMO channel;asymmetric link strengths;block-fading noncoherent multiple input multiple output channels;asymmetric distributions;optimal signaling structure;asymmetric MIMO channel;single input multiple output channel;asymmetric case;SIMO channel;arbitrary link strengths;multiple input single output channel;direct links;cross links;gDoF gain;degrees of freedom optimal;MxM MIMO channel;MIMO communication;Signal to noise ratio;Fading channels;Transmitting antennas;Coherence;MISO communication;Degrees of freedom (DoF);multiple antennas;time-varying channels;noncoherent communication},\n month = {July},\n number = {7},\n pages = {4431-4448},\n tags = {journal,IT,ANIT,WiNetnew,NCWN,WNIF},\n title = {Generalized Degrees Freedom of Noncoherent MIMO Channels With Asymmetric Link Strengths},\n type = {2},\n volume = {66},\n year = {2020}\n}\n\n
\n
\n\n\n
\n We study the generalized degrees of freedom (gDoF) of block-fading noncoherent multiple input multiple output (MIMO) channels with asymmetric distributions of link strengths and a coherence time of T symbol durations. We derive the optimal signaling structure for communication for the asymmetric MIMO channel, which is distinct from that for the MIMO channel with independent and identically distributed (i.i.d.) links. We extend the existing results for the single input multiple output (SIMO) channel with i.i.d. links to the asymmetric case, proving that selecting the statistically best antenna is gDoFoptimal. Using the gDoF result for the SIMO channel, we prove that for T = 1, the gDoF is zero for MIMO channels with arbitrary link strengths. We show that selecting the statistically best antenna is gDoF-optimal for the multiple input single output (MISO) channel. We also derive the gDoF for the 2 x 2 MIMO channel with different exponents in the direct and cross links. In this setting, we show that it is always necessary to use both the antennas to achieve the gDoF, in contrast to the results for the 2 x 2 MIMO channel with i.i.d. links. We show that having weaker crosslinks, gives gDoF gain compared to the case with i.i.d. links. For the noncoherent MIMO channel with i.i.d. links, the traditional method of training each transmit antenna independently is degrees of freedom (DoF) optimal, whereas we observe that for the asymmetric 2 x 2 MIMO channel, the traditional training is not gDoF-optimal. We extend this observation to a larger MxM MIMO channel by demonstrating a strategy that can achieve larger gDoF than a traditional trainingbased method.\n
\n\n\n
\n\n\n
\n
\n\n \n \n J. Sebastian; and S. Diggavi.\n\n\n \n \n \n \n \n Generalized Degrees of Freedom of Noncoherent Diamond Networks.\n \n \n \n \n\n\n \n\n\n\n
IEEE Transactions on Information Theory, 66(8): 5228-5260. Aug 2020.\n
\n\n
\n\n
\n\n
\n\n \n \n arxiv\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 3 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{9046857,\n abstract = {We study the generalized degrees of freedom (gDoF) of the noncoherent diamond (parallel relay) wireless network with asymmetric distributions of link strengths. We use the noncoherent block-fading model introduced by Marzetta and Hochwald, where no channel state information is available at the transmitters or at the receivers and the channels remain constant for a coherence time of T symbol durations. We first derive an upper bound for the capacity of this channel and then derive the optimal structure for the solution of the upper bound optimization problem. Using the optimal structure, we solve the upper bound optimization problem in terms of its gDoF. Using insights from our upper bound signaling solution, we devise an achievability strategy based on a novel scheme that we call train-scale quantize-map-forward (TS-QMF). This scheme uses training in the links from the source to the relays, scaling and quantizing at the relays combined with nontraining-based schemes. We show the optimality of this scheme by comparing it to the upper bound in terms of the gDoF. In noncoherent point-to-point multiple-input-multiple-output (MIMO) channels, where the fading realization is unknown to the transmitter and the receiver, an important tradeoff between communication and channel learning was revealed by Zheng and Tse, by demonstrating that not all the available antennas might be used, as it is suboptimal to learn all their channel parameters. Our results in this paper for the diamond network demonstrate that in certain regimes of relative channel strengths, the gDoF-optimal scheme uses a subnetwork, demonstrating a similar tradeoff between channel learning and communication. In some regimes, it is gDoF-optimal to do relay selection, i.e., use a part of the network. In the other regimes, even when it is essential to use the entire network, it is suboptimal to learn the channel states for all the links in the network, i.e., traditional training-based schemes are suboptimal in these regimes.},\n author = {J. {Sebastian} and S. {Diggavi}},\n doi = {10.1109/TIT.2020.2983169},\n issn = {1557-9654},\n journal = {IEEE Transactions on Information Theory},\n keywords = {channel capacity;fading channels;learning (artificial intelligence);MIMO communication;optimisation;quantisation (signal);radio receivers;radio transmitters;relay networks (telecommunication);telecommunication computing;noncoherent point-to-point MIMO channel;noncoherent point-to-point multiple-input-multiple-output channel;TS-QMF;bound signaling solution;bound optimization problem;channel capacity;Hochwald;Marzetta;block-fading model;asymmetric distribution;noncoherent diamond wireless network;generalized degrees of freedom;noncoherent diamond network;gDoF-optimal scheme;train-scale quantize-map-forward;channel state information;Signal to noise ratio;Relays;Diamond;MIMO communication;Upper bound;Fading channels;Wireless networks;Noncoherent communication;degrees of freedom (DoF);relay channels;diamond network;time-varying channels},\n month = {Aug},\n number = {8},\n pages = {5228-5260},\n tags = {journal,IT,ANIT,WiNetnew,NCWN,WNIF},\n title = {Generalized Degrees of Freedom of Noncoherent Diamond Networks},\n type = {2},\n url_arxiv = {https://arxiv.org/abs/1802.02667},\n volume = {66},\n year = {2020}\n}\n\n
\n
\n\n\n
\n We study the generalized degrees of freedom (gDoF) of the noncoherent diamond (parallel relay) wireless network with asymmetric distributions of link strengths. We use the noncoherent block-fading model introduced by Marzetta and Hochwald, where no channel state information is available at the transmitters or at the receivers and the channels remain constant for a coherence time of T symbol durations. We first derive an upper bound for the capacity of this channel and then derive the optimal structure for the solution of the upper bound optimization problem. Using the optimal structure, we solve the upper bound optimization problem in terms of its gDoF. Using insights from our upper bound signaling solution, we devise an achievability strategy based on a novel scheme that we call train-scale quantize-map-forward (TS-QMF). This scheme uses training in the links from the source to the relays, scaling and quantizing at the relays combined with nontraining-based schemes. We show the optimality of this scheme by comparing it to the upper bound in terms of the gDoF. In noncoherent point-to-point multiple-input-multiple-output (MIMO) channels, where the fading realization is unknown to the transmitter and the receiver, an important tradeoff between communication and channel learning was revealed by Zheng and Tse, by demonstrating that not all the available antennas might be used, as it is suboptimal to learn all their channel parameters. Our results in this paper for the diamond network demonstrate that in certain regimes of relative channel strengths, the gDoF-optimal scheme uses a subnetwork, demonstrating a similar tradeoff between channel learning and communication. In some regimes, it is gDoF-optimal to do relay selection, i.e., use a part of the network. In the other regimes, even when it is essential to use the entire network, it is suboptimal to learn the channel states for all the links in the network, i.e., traditional training-based schemes are suboptimal in these regimes.\n
\n\n\n
\n\n\n\n\n\n