The distance to uncontrollability for a linear control system is the (in 2-norm) nearest uncontrollable system. We present an algorithm based on methods of Gu and Burke-Lewis-Overton that estimates any prescribed accuracy. new method requires O(n4 ) operations average, which improvement over previous have complexity O(n6 ), where n order Numerical experiments indicate reliable in practice.

We consider the symmetric Gaussian interference channel where two users try to enhance their secrecy rates in a cooperative manner.Artificial noise is introduced along with useful information.We derive power control and artificial parameter for kinds of optimal points, max-min point single user point.It shown that there exists critical value Pc constraint, below which an on rate region, above time-sharing between points achieves larger pairs.It also can help enlarge particular point.

Lattice codes used under the compute-and-forward paradigm suggest an alternative strategy for standard Gaussian multiple-access channel (MAC): receiver successively decodes integer linear combinations of messages until it can invert and recover all messages. In this paper, a technique called compute-forward multiple access (CFMA) is proposed analyzed. For two-user MAC, shown that without time-sharing, entire capacity region be attained using CFMA with single-user decoder as soon...

Hardware-limited task-based quantization is a new design paradigm for data acquisition systems equipped with serial scalar analog-to-digital converters using small number of bits. By taking into account the underlying system task, quantizers can efficiently recover desired parameters from low-bit quantized observation. Current and analysis frameworks hardware-limited are only applicable to inputs bounded support uniform non-subtractive dithering. Here, we propose framework based on...

Transfer learning is a machine paradigm where knowledge from one problem utilized to solve new but related problem. While conceivable that task could help task, if not executed properly, transfer algorithms can impair the performance instead of improving it – commonly known as negative transfer. In this paper, we use parametric statistical model study Bayesian perspective. Specifically, three variants problems, instantaneous, online, and time-variant learning. We define an appropriate...

We present a modified compute-and-forward scheme which utilizes Channel State Information at the Transmitters (CSIT) in natural way. The allows different users to have coding rates, and use CSIT achieve larger rate region. This idea is applicable all systems technique can be arbitrarily better than regular some settings.

Semi-supervised learning (SSL) aims to train a machine (ML) model using both labeled and unlabeled data. While the data have been used in various ways improve prediction accuracy, reason why could help is not fully understood. One interesting promising direction understand SSL from causal perspective. In light of independent mechanisms (ICM) principle, can be helpful when label causes features but vice versa. However, relations between labels complex real world applications. this article, we...

Building on the previous work of Lee et al. [2] and Ferdinand [3] coded computation, we propose a sequential approximation framework for solving optimization problems in distributed manner. In computation system, latency caused by individual processors ("stragglers") usually causes significant delay overall process. The proposed method is powered scheme, which designed specifically systems with stragglers. This scheme has desirable property that user guaranteed to receive useful...

We present a modified compute-and-forward scheme which utilizes Channel State Information at the Transmitters (CSIT) in natural way. The allows different users to have coding rates, and use CSIT achieve larger rate region. This idea is applicable all systems technique can be arbitrarily better than regular some settings.

Transfer learning, or domain adaptation, is concerned with machine learning problems in which training and testing data come from possibly different distributions (denoted as μ μ', respectively). In this work, we give an informationtheoretic analysis on the generalization error excess risk of transfer algorithms, following a line work initiated by Russo Zhou. Our results suggest, perhaps expected, that Kullback-Leibler (KL) divergence D(μ||μ') plays important role characterizing settings...

We present a practical strategy that aims to attain rate points on the dominant face of multiple access channel capacity using standard low complexity decoder. This technique is built upon recent theoretical developments Zhu and Gastpar compute-forward which achieves sequential illustrate this with off-the-shelf LDPC codes. In first stage decoding, receiver recovers linear combination transmitted codewords sum-product algorithm (SPA). second stage, by recovered sum-of-codewords as side...

Transfer learning is a machine paradigm where the knowledge from one task utilized to resolve problem in related task. On hand, it conceivable that could be useful for solving problem. other also recognized if not executed properly, transfer algorithms fact impair performance instead of improving - commonly known as negative transfer. In this paper, we study online problems source samples are given an off-line way while target arrive sequentially. We define expected regret problem, and...

Inspired by the compute-and-forward scheme from Nazer and Gastpar, a novel multiple-access introduced Zhu Gastpar makes use of nested lattice codes sequential decoding linear combinations codewords to recover individual messages. This strategy, coined compute-forward multiple access (CFMA), provably achieves points on dominant face capacity region while circumventing need time sharing or rate splitting. For two-user channel (MAC), we propose practical procedure design suitable off-the-shelf...

Real-time data-driven optimization and control problems over networks may require sensitive information of participating users to calculate solutions decision variables, such as in traffic or energy systems. Adversaries with access coordination signals potentially decode on individual put user privacy at risk. We develop local differential privacy, which is a strong notion that guarantees regardless any auxiliary an adversary have, for larger family convex distributed problems. The mechanism...

We analyze the symbol error probability (SEP) of $M$-ary pulse amplitude modulation ($M$-PAM) receivers equipped with optimal low-resolution quantizers. first show that optimum detector can be reduced to a simple decision rule. Using this simplification, an exact SEP expression for quantized $M$-PAM is obtained when Nakagami-$m$ fading channel considered. The derived enables optimization quantizer and/or constellation under minimum criterion. Our analysis quantization equidistant receiver...

Due to mutual interference between users, power allocation problems in wireless networks are often non-convex and computationally challenging. Graph neural (GNNs) have recently emerged as a promising approach tackling these an that exploits the underlying topology of networks. In this paper, we propose novel graph representation method for include full-duplex (FD) nodes. We then design corresponding FD Neural Network (F-GNN) with aim allocating transmit powers maximise network throughput....

Transfer learning, or domain adaptation, is concerned with machine learning problems in which training and testing data come from possibly different distributions (denoted as $\mu$ $\mu'$, respectively). In this work, we give an information-theoretic analysis on the generalization error excess risk of transfer algorithms, following a line work initiated by Russo Zhou. Our results suggest, perhaps expected, that Kullback-Leibler (KL) divergence $D(mu||mu')$ plays important role characterizing...

The usefulness of lattice codes is investigated for two-user Gaussian interference channels (IC). A coding scheme based on the compute-and-forward technique shown to achieve capacity region IC under strong interference. proposed uses single-user decoders whereas conventional multi-user simultaneous decoding. same applicable Z-interference channel. lattice-based also devised state-dependent with state sequence non-causally known transmitters. establishes new achievable rate regions, which can...