This paper investigates sparse signal recovery based on expectation propagation (EP) from unitarily invariant measurements. A rigorous analysis is presented for the state evolution (SE) of an EP-based message-passing algorithm in large system limit, where both input and output dimensions tend to infinity at identical speed. The main result justification SE formula conjectured by Ma Ping.

Signal recovery from unitarily invariant measurements is investigated in this paper. A message-passing algorithm formulated on the basis of expectation propagation (EP). rigorous analysis presented for dynamics large system limit, where both input and output dimensions tend to infinity while compression rate kept constant. The main result justification state evolution (SE) equations conjectured by Ma Ping. This implies that EP-based achieves Bayes-optimal performance was originally derived...

This paper proposes Bayes-optimal convolutional approximate message-passing (CAMP) for signal recovery in compressed sensing. CAMP uses the same low-complexity matched filter (MF) interference suppression as (AMP). To improve convergence property of AMP ill-conditioned sensing matrices, so-called Onsager correction term is replaced by a convolution all preceding messages. The tap coefficients are determined so to realize asymptotic Gaussianity estimation errors via state evolution (SE) under...

Kudekar et al. proved that the belief-propagation (BP) performance for low-density parity check (LDPC) codes can be boosted up to maximum-a-posteriori (MAP) by spatial coupling. In this paper, coupling is applied sparsely-spread code-division multiple-access (CDMA) systems improve of iterative multiuser detection based on BP. Two receivers BP are considered: One receiver exact and other an approximate with Gaussian approximation. The two evaluated via density evolution (DE) in dense limit...

This paper proposes decentralized generalized approximate message-passing (D-GAMP) for compressed sensing in a tree-structured network. In contrast to conventional GAMP that needs central node gather all measurements on the network, D-GAMP utilizes consensus propagation via between adjacent nodes achieve same performance as centralized GAMP. The Onsager correction is designed rigorous state evolution so asymptotic Gaussianity of estimation errors guaranteed each iteration propagation. By...

Kudekar et al. proved that the belief-propagation (BP) threshold for low-density parity-check codes can be boosted up to maximum-a-posteriori (MAP) by spatial coupling. In this paper, coupling is applied randomly-spread code-division multiple-access (CDMA) systems in order improve performance of BP-based multiuser detection (MUD). Spatially-coupled CDMA regarded as multi-code with two transmission phases. The large-system analysis shows BP performance, while there a gap between and...

Training-based transmission over Rayleigh block-fading multiple-input multiple-output (MIMO) channels is investigated.As a training method combination of pilot-assisted scheme and biased signaling considered.The achievable rates successive decoding (SD) receivers based on the linear minimum mean-squared error (LMMSE) channel estimation are analyzed in large-system limit, by using replica under assumption symmetry.It shown that negligible pilot information best terms SD limit.The obtained...

Orthogonal/vector approximate message-passing (AMP) is a powerful (MP) algorithm for signal reconstruction in compressed sensing. This paper proves the convergence of Bayes-optimal orthogonal/vector AMP large system limit. The proof strategy based on novel long-memory (LM) MP approach: A first step construction LM-MP that guaranteed to converge systematically. second large-system analysis via an existing framework state evolution. third prove evolution recursions new statistical...

Kudekar et al. proved an interesting result in low-density parity-check (LDPC) convolutional codes: The belief-propagation (BP) threshold is boosted to the maximum-a-posteriori (MAP) by spatial coupling. Furthermore, authors showed that BP for code-division multiple-access (CDMA) systems improved up optimal one via In this letter, a phenomenological model elucidating essence of these phenomenon, called improvement, proposed. main implies improvement occurs spatially-coupled general graphical models.

This paper presents a large-system analysis of the performance joint channel estimation, multiuser detection, and per-user decoding (CE-MUDD) for randomly-spread multiple-input multiple-output (MIMO) direct-sequence code-division multiple-access (DS-CDMA) systems. A suboptimal receiver based on successive in conjunction with linear minimum mean-squared error (LMMSE) estimation is investigated. The replica method, developed statistical mechanics, used to evaluate limit, where number users...

This paper investigates iterative detection based on expectation propagation (EP) in overloaded multiple-input multiple-output systems. The conjugate gradient (CG) method is utilized to reduce the per-iteration complexity of EP detection. Under assumption unitarily invariant channel matrices, dynamics CG (EP-CG) analyzed large system limit, where transmit and receive dimensions tend infinity at an identical rate. main result a rigorous derivation state evolution for EP-CG.

This paper presents a unified framework to understand the dynamics of message-passing algorithms in compressed sensing. State evolution is rigorously analyzed for general error model that contains approximate (AMP), as well orthogonal AMP. As byproduct, AMP proved converge asymptotically if sensing matrix orthogonally invariant and moment sequence its asymptotic singular-value distribution coincide with Marčhenko-Pastur up order at most twice large maximum number iterations.

We analyze a MIMO DS-CDMA channel with general multiuser detector including nonlinear detector, using the replica method. In many-user, limit is decoupled into bank of single-user SIMO Gaussian channels if spatial spreading scheme employed. On other hand, it not The spectral efficiency comparable that an optimal space-time block code without scheme. case QPSK data modulation MMSE shows {\it waterfall} behavior and very close to corresponding sum capacity when system load just below...

To investigate if uroflowmetry (UFM) curve patterns could differentiate between detrusor underactivity (DU) and bladder outlet obstruction (BOO).A hundred consecutive data sets of male patients who were evaluated using UFM invasive urodynamics (pressure flow study) retrospectively collected. DU BOO diagnosed according to the contractility index index. The with two or more notches was defined as sawtooth pattern, interrupted pattern several curves interruptions reducing zero noted. We also...

This paper proves the convergence of Bayes-optimal orthogonal/vector approximate message-passing (AMP) to a fixed point in large system limit. The proof is based on long-memory (LM) (MP) that guaranteed converge systematically. dynamics LM-MP analyzed via an existing state evolution framework. obtained recursions are proved converge. AMP by confirming exact reduction those for AMP.

Spatially-coupled (SC) codes are constructed by coupling many regular low-density parity-check in a chain. The decoding chain of SC aborts when facing burst erasures. This problem cannot be overcome increasing the length. In this paper, we introduce multi-dimensional (MD) to circumvent it. Numerical results show that two-dimensional more robust against erasures than one-dimensional codes. Furthermore, consider designing multidimensional with smaller rateloss.

This letter proposes a novel message-passing algorithm for signal recovery in compressed sensing. The proposed solves the disadvantages of approximate (AMP) and orthogonal/vector AMP, realizes their advantages. AMP converges only limited class sensing matrices while it has low complexity. Orthogonal/vector requires high-complexity matrix inversion is applicable wide matrices. key feature so-called Onsager correction via convolution messages all preceding iterations conventional algorithms...

This paper investigates sparse signal recovery based on expectation propagation (EP) from unitarily invariant measurements. A rigorous analysis is presented for the state evolution (SE) of an EP-based message-passing algorithm in large system limit, where both input and output dimensions tend to infinity at identical speed. The main result justification SE formula conjectured by Ma Ping.

Signal recovery from unitarily invariant measurements is investigated in this paper. A message-passing algorithm formulated on the basis of expectation propagation (EP). rigorous analysis presented for dynamics large system limit, where both input and output dimensions tend to infinity while compression rate kept constant. The main result justification state evolution (SE) equations conjectured by Ma Ping. This implies that EP-based achieves Bayes-optimal performance was originally derived...

Kudekar et al. proved that the belief-propagation (BP) threshold for low-density parity-check codes can be boosted up to maximum-a-posteriori (MAP) by spatial coupling. In this paper, coupling is applied randomly-spread code-division multiple-access (CDMA) systems in order improve performance of BP-based multiuser detection (MUD). Spatially-coupled CDMA regarded as multi-code with two transmission phases. The large-system analysis shows BP performance, while there a gap between and...

Spatially coupled (SC) interleaving is proposed to improve the performance of iterative multiuser detection and decoding (MUDD) for quasi-static fading multiple-input multiple-output systems. The linear minimum mean-squared error (LMMSE) demodulator used reduce complexity avoid propagation. Furthermore, sliding window MUDD circumvent an increase latency due SC interleaving. Theoretical numerical analyses show that can LMMSE regular low-density parity-check codes.