Logs of large-scale cloud systems record diverse system events, ranging from routine statuses to critical errors. As the fundamental step of automated log analysis, log parsing is to transform unstructured logs into structured data for easier management and analysis. However, existing syntax-based and deep learning-based parsers struggle with complex real-world logs. Recent parsers based on large language models (LLMs) achieve higher accuracy, but they typically rely on online APIs (e.g., ChatGPT), raising privacy concerns and suffering from network latency. Moreover, with the rise of artificial intelligence for IT operations (AIOps), traditional parsers that focus on syntax-level templates fail to capture the semantics of dynamic log parameters, limiting their usefulness for downstream tasks. These challenges highlight the need for semantic log parsing that goes beyond template extraction to understand parameter semantics.This paper presents SemanticLog, an effective and efficient semantic log parser powered by open-source LLMs. SemanticLog adapts the structure of LLMs to the log parsing task, leveraging their rich knowledge while safeguarding log data privacy. It first extracts informative feature representations from log data, then refines them through fine-grained semantic perception to enable accurate template and parameter extraction together with semantic category prediction. To boost scalability, SemanticLog introduces the EffiParsing tree for faster inference on large-scale logs. Extensive experiments on the LogHub-2.0 dataset show that SemanticLog significantly outperforms the state-of-the-art log parsers in terms of accuracy. Moreover, it also surpasses existing LLM-based parsers in efficiency while showcasing advanced semantic parsing capability. Notably, SemanticLog employs much smaller open-source LLMs compared to existing LLM-based parsers (mainly based on ChatGPT), while maintaining better capability of log data privacy protection.

Distributed traces contain valuable information but are often massive in volume, posing a core challenge in tracing framework design: balancing the tradeoff between preserving essential trace information and reducing trace volume. To address this tradeoff, previous approaches typically used a '1 or 0' sampling strategy: retaining sampled traces while completely discarding unsampled ones. However, based on an empirical study on real-world production traces, we discover that the '1 or 0' strategy actually fails to effectively balance this tradeoff.

Multivariate Time Series Forecasting (MTSF) is a critical task in various domains, and Large Language Models (LLMs) for MTSF have recently received considerable attention. Despite significant progress in large-scale time series models, particularly in fine-tuning pre-trained LLMs for MTSF, there are still limitations with existing works. First, multivariate time series (MTS) are often handled as multiple independent univariate inputs and processed separately across different variates, which neglects the dependencies between the variates. Second, most existing approaches employ Mean Squared Error (MSE) as loss function, which evaluates error at each time point separately, ignoring long-term dependencies. To address these limitations, this paper explores inter-variate and long-term dependencies to boost MTSF performance. We propose a temporal channel adapter to capture inter-variate relationships, and introduce a post-constraint module to model correlations between consecutive time points. Extensive experiments on benchmark datasets show that our method achieves state-of-the-art performance across diverse datasets and prediction horizons.

Logs record essential information about system operations and serve as a critical source for anomaly detection, which has generated growing research interest. Utilizing large language models (LLMs) within a retrieval-augmented generation (RAG) framework for log-based anomaly detection is an effective approach due to its strong generalization capabilities and efficient few-shot performance. However, the effectiveness of this method hinges on the quality of the knowledge source, which can be impacted by noise and changes within the software systems. Facing these problems, in this paper, we propose a novel log-based anomaly detection method named EagerLog, employing active learning to choose the logs for humans to label, thereby adding them to the knowledge source, thus enhancing the knowledge source and maintaining its quality. Our experiments on three open datasets (BGL, Thunderbird, Zookeeper) and one industrial dataset demonstrate that EagerLog can achieve 93.65% F1 score with approximately 10 labeled log sequences, surpassing existing methods by 15.32%.

Accurately diagnosing the fault that causes the failure is crucial for maintaining the reliability of a microservice system after a failure occurs. Mainstream fault diagnosis approaches are data-driven and mainly rely on three modalities of runtime data: traces, logs, and metrics. Diagnosing faults with multiple modalities of data in microservice systems has been a clear trend in recent years because different types of faults and corresponding failures tend to manifest in data of various modalities. Accurately diagnosing faults by fully leveraging multiple modalities of data is confronted with two challenges: 1) how to minimize information loss when extracting features for data of each modality; 2) how to correctly capture and utilize the relationships among data of different modalities. To address these challenges, we propose FAMOS, a Fault diagnosis Approach for MicrOservice Systems through effective multi-modal data fusion. On the one hand, FAMOS employs independent feature extractors to preserve the intrinsic features for each modality. On the other hand, FAMOS introduces a new Gaussian-attention mechanism to accurately correlate data of different modalities and then captures the inter-modality relationship with a crossattention mechanism. We evaluated FAMOS on two datasets constructed by injecting comprehensive and abundant faults into an open-source microservice system and a real-world industrial microservice system. Experimental results demonstrate the FAMOS's effectiveness in fault diagnosis, achieving significant improvements in F1 scores compared to state-of-the-art (SOTA) methods, with an increase of 20.33 %.

Logs are ubiquitous digital footprints, playing an indispensable role in system diagnostics, security analysis, and performance optimization. The extraction of actionable insights from logs is critically dependent on the log parsing process, which converts raw logs into structured formats for downstream analysis. Yet, the complexities of contemporary systems and the dynamic nature of logs pose significant challenges to existing automatic parsing techniques. The emergence of Large Language Models (LLM) offers new horizons. With their expansive knowledge and contextual prowess, LLMs have been transformative across diverse applications. Building on this, we introduce LogParser-LLM, a novel log parser integrated with LLM capabilities. This union seamlessly blends semantic insights with statistical nuances, obviating the need for hyper-parameter tuning and labeled training data, while ensuring rapid adaptability through online parsing. Further deepening our exploration, we address the intricate challenge of parsing granularity, proposing a new metric and integrating human interactions to allow users to calibrate granularity to their specific needs. Our method's efficacy is empirically demonstrated through evaluations on the Loghub-2k and the large-scale LogPub benchmark. In evaluations on the LogPub benchmark, involving an average of 3.6 million logs per dataset across 14 datasets, our LogParser-LLM requires only 272.5 LLM invocations on average, achieving a 90.6% F1 score for grouping accuracy and an 81.1% for parsing accuracy. These results demonstrate the method's high efficiency and accuracy, outperforming current state-of-the-art log parsers, including pattern-based, neural network-based, and existing LLM-enhanced approaches.

Cloud microservices are being scaled up due to the rising demand for new features and the convenience of cloud-native technologies. However, the growing scale of microservices complicates the remote procedure call (RPC) dependency graph, exacerbates the tail-of-scale effect, and makes many of the empirical rules for detecting the root cause of end-to-end performance issues unreliable. Additionally, existing open-source microservice benchmarks are too small to evaluate performance debugging algorithms at a production-scale with hundreds or even thousands of services and RPCs.

Lenstra-Lenstra-Lovász (LLL) lattice reduction technique has been applied to multiple-input multiple-output (MIMO) detectors to collect full diversity while enjoying low complexity. However, the original LLL algorithm has variable complexity, which is not desirable in hardware implementation. To solve this problem, some fixed-complexity LLL (fcLLL) algorithms have recently been proposed by using the fixed-column traverse strategy with limited number of LLL iterations. The existing fcLLL algorithms are designed to process each column with equal priority, which is not optimized in terms of error performance and complexity. In this paper, we propose an enhanced fcLLL algorithm with a novel column traverse strategy by allocating priorities to columns based on the characteristics of LLL and MIMO detection. In addition, we propose an improved termination criterion without sacrificing the error performance in the proposed fcLLL algorithm. Simulations show that our proposed fcLLL algorithm converges faster than LLL and existing fcLLL algorithms, and yields better error performance than the LLL and existing fcLLL algorithms when the maximum number of LLL iterations is fixed. Furthermore, in large MIMO systems, our proposed fcLLL algorithm exhibits significant complexity advantage, saving about 90% LLL iterations in average compared to the existing fcLLL algorithms for a 128× 128 MIMO system with 64-QAM.

Most existing differential receivers for ultra-wideband (UWB) communications employ Gaussian approximation for multiple access interference (MAI). However, for a system with strong interference caused by a small number of active users, significant performance degradation is found for differential UWB receivers due to the impreciseness of Gaussian approximation on impulsive MAI. In this paper, we propose new differential UWB receivers based on the generalized Gaussian (GG) distribution, which subsumes Gaussian distribution and Laplace distribution as special cases. Numerical results show that the GG distribution approximates MAI well. In addition, we show that GG-based multiple-symbol differential receivers can be formulated as the same form as the conventional multiple-symbol differential receivers, which can be efficiently solved using existing algorithms. Simulations are conducted to demonstrate the superior performance of the proposed receivers to that of the existing differential UWB receivers when the interference caused by a small number of active users is strong.

With the aim of extending the coverage and improving the performance of impulse radio ultra-wideband (UWB) systems, this paper focuses on developing a novel single differential encoded decode and forward (DF) non-cooperative relaying scheme (NCR). To favor simple receiver structures, differential noncoherent detection is employed which enables effective energy capture without any channel estimation. Putting emphasis on the general case of multi-hop relaying, we illustrate an original algorithm for the joint power allocation and path selection (JPAPS), minimizing an approximate expression of the overall bit error rate (BER). In particular, after deriving a closed-form power allocation strategy, the optimal path selection is reduced to a shortest path problem on a connected graph, which can be solved without any topology information with complexity O(N3), N being the number of available relays of the network. An approximate scheme is also presented, which reduces the complexity to O(N2) while showing a negligible performance loss, and for benchmarking purposes, an exhaustive-search based multi-hop DF cooperative strategy is derived. Simulation results for various network setups corroborate the effectiveness of the proposed low-complexity JPAPS algorithm, which favorably compares to existing AF and DF relaying methods.

Multiple-input multiple-output (MIMO) techniques provide high data rates but the optimal maximum likelihood (ML) detector exhibits high complexity. Recently lattice reduction (LR) aided detectors have been proposed to achieve near-ML performance with low complexity. In this paper, we develop a LR-aided complex K-best algorithm which reduces the complexity of the existing sphere decoding based K-best algorithm. Then we provide the fixed-point design of the LR-aided K-best MIMO receiver for both coded and uncoded systems. The architecture selection of each sub-module is developed and a simulation-based wordlength optimization procedure is proposed. Simulations show that the fixed-point results can keep bit error rate degradation within 0.2dB under 8 × 8 256-QAM MIMO systems.

Large multi-input multi-output (MIMO) systems with tens or hundreds of antennas have shown great potential for next generation of wireless communications to support high spectral efficiencies. However, due to the non-deterministic polynomial hard nature of MIMO detection, large MIMO systems impose stringent requirements on the design of reliable and computationally efficient detectors. Recently, lattice reduction (LR) techniques have been applied to improve the performance of low-complexity detectors for MIMO systems without increasing the complexity dramatically. Most existing LR algorithms are designed to improve the orthogonality of channel matrices, which is not directly related to the error performance. In this paper, we propose element-based lattice reduction (ELR) algorithms that reduce the diagonal elements of the noise covariance matrix of linear detectors and thus enhance the asymptotic performance of linear detectors. The general goal is formulated as solving a "shortest longest vector reduction" or a stronger version, "shortest longest basis reduction," both of which require high complexity to find the optimal solution. Our proposed ELR algorithms find sub-optimal solutions to the reductions with low complexity and high performance. The fundamental properties of the ELR algorithms are investigated. Simulations show that the proposed ELR-aided detectors yield better error performance than the existing low-complexity detectors for large MIMO systems while maintaining lower complexity.

Lattice-reduction (LR)-aided linear detectors (LDs) have shown great potentials for wireless communications due to their low complexity and high performance. However, most of the existing LR algorithms do not directly aim at minimizing the asymptotic error performance of LDs, which is dominated by the shortest longest vector (SLV) in the dual space. To find sub-optimal solutions to the SLV reduction, element-based lattice reduction (ELR) algorithms were recently proposed by performing column-addition operations. In this paper, we propose improved ELR algorithms (called ELR+) by performing generalized column-addition operations. We find that the problem that minimizes a basis vector by a generalized column-addition operation can be formulated as a closest vector problem (CVP). By solving the CVP that minimizes the longest basis vector for each basis update, the proposed ELR+algorithms find sub-optimal solutions to the SLV reduction problem with high performance. Simulations illustrate that the proposed ELR+algorithms show superior performance relative to the state-of-the-art LRs for linear detection, including Korkin-Zolotarev reductions.

Code-multiplexed transmitted reference (CM-TR) and code-shifted reference (CSR) have recently drawn attention in the field of ultra-wideband communications mainly because they enable noncoherent detection without requiring either a delay component, as in transmitted reference, or an analog carrier, as in frequency-shifted reference, to separate the reference and data-modulated signals at the receiver. In this paper, we propose a generalized code-multiplexing (GCM) system based on the formulation of a constrained mixed-integer optimization problem. The GCM extends the concept of CM-TR and CSR while retaining their simple receiver structure, even offering better bit-error-rate performance and a higher data rate in the sense that more data symbols can be embedded in each transmitted block. The GCM framework is further extended to the cases when peak power constraint is considered and when inter-frame interference exists, as typically occurs in high data-rate transmissions. Numerical simulations performed over demanding wireless environments corroborate the effectiveness of the proposed approach.

This paper proposes a novel cooperative approach for two-hop amplify-and-forward (A&F) relaying that exploits both the signal forwarded by the relay and the one directly transmitted by the source in impulse-radio ultra-wideband (IR-UWB) systems. Specifically, we focus on a non-coherent setup employing a double-differential encoding scheme at the source node and a single differential demodulation at the relay and destination. The log-likelihood ratio based decision rule is derived at the destination node. A semi-analytical power allocation strategy is presented by evaluating a closed-form expression for the effective signal to noise ratio (SNR) at the destination, which is maximized by exhaustive search. Numerical simulations show that the proposed system outperforms both the direct transmission with single differential encoding and the non-cooperative multi-hop approach in different scenarios.

Recently, multi-symbol based transmitted reference system has drawn attention for UWB communications because of its high performance without estimating the channel explicitly. The maximum-likelihood (ML) detector for multi-symbol transmitted reference system has been proposed to jointly detect the multi-symbol and yields a considerable improvement compared with conventional single symbol detector. However, the high computational complexity incurred by ML detectors hinders their applications to practical systems. In this paper, we propose a polynomial-time (in worse case) approximation near-ML detector based on semi-definite programming (SDP). Simulation results demonstrate that the proposed SDP detector provides almost the same BER performance as the ML detectors and is robust to the multi-access interference.