With some specific characteristics such as elastics and scalability, cloud computing has become the most promising technology for online business nowadays. However, how to efficiently perform real-time job scheduling in cloud still poses significant challenges. The reason is that those jobs are highly dynamic and complex, and it is always hard to allocate them to computing resources in an optimal way, such as to meet the requirements from both service providers and users. In recent years, various works demonstrate that deep reinforcement learning (DRL) can handle real-time cloud jobs well in scheduling. However, to our knowledge, none of them has ever considered extra optimization opportunities for the allocated jobs in their scheduling frameworks. Given this fact, in this work, we introduce a novel DRL-based preemptive method for further improve the performance of the current studies. Specifically, we try to improve the training of scheduling policy with effective job preemptive mechanisms, and on that basis to optimize job execution cost while meeting users' expected response time. We introduce the detailed design of our method, and our evaluations demonstrate that our approach can achieve better performance than other scheduling algorithms under different real-time workloads, including the DRL approach.

Wolf Pack Algorithm (WPA) is a swarm intelligence algorithm that simulates the food searching process of wolves. It is widely used in various engineering optimization problems due to its global convergence and computational robustness. However, the algorithm has some weaknesses such as low convergence speed and easily falling into local optimum. To tackle the problems, we introduce an improved approach called OGL-WPA in this work, based on the employments ofOpposition-based learning andGenetic algorithm withLevy’s flight. Specifically, in OGL-WPA, the population of wolves is initialized by opposition-based learning to maintain the diversity of the initial population during global search. Meanwhile, the leader wolf is selected by genetic algorithm to avoid falling into local optimum and the round-up behavior is optimized by Levy’s flight to coordinate the global exploration and local development capabilities. We present the detailed design of our algorithm and compare it with some other nature-inspired metaheuristic algorithms using various classical test functions. The experimental results show that the proposed algorithm has better global and local search capability, especially in the presence of multi-peak and high-dimensional functions.

Due to their important applications in theory and practice, linear complementary dual (LCD) codes and self-orthogonal codes have received much attention in the last decade. The objective of this paper is to extend a recent construction of binary LCD codes and self-orthogonal codes to the general $ p $-ary case, where $ p $ is an odd prime. Based on the extended construction, several classes of $ p $-ary linear codes are obtained. The characterizations of these linear codes to be LCD or self-orthogonal are derived. The duals of these linear codes are also studied. It turns out that the proposed linear codes are optimal in many cases in the sense that their parameters meet certain bounds on linear codes. The weight distributions of these linear codes are settled.

Aiming at the shortcomings of long completion time and high consumption cost of cloud computing batch task scheduling in IoT, an Improved Chicken Swarm Optimization Algorithm (ICSO) for task scheduling in cloud computing scenarios is proposed. Specifically, in order to solve the problems of slow convergence and falling into local optimum of the chicken swarm optimization algorithm, we adopt the nonlinear decreasing technique of the rooster and the weighting technique of the hen, optimize the following coefficients of the chicks, and apply ICSO to cloud computing task scheduling. In simulation experiments, we conducted a large number of experiments using four standard benchmark functions with different number of tasks and the results show that ICSO algorithm reduces 25.8%, 9.3%, 8.8%, 7.5% in small task time compared to CSO, DCSO, GCSO, ABCSO in large task time by 30.8%, 8.3%, 7.8%, 6.3%, 11.8%, 10.3%, 8.8%, 7.5% savings in small task cost and 25.8%, 11.2%, 10.8%, 9.3% savings in large task cost. This method effectively reduces task scheduling time and cost consumption. Meanwhile, we tested it in combination with an IoT-based cloud platform and achieved very satisfying Results.

Local codes are a special kind of error-correcting codes. Locally correctable codes (LCCs) are one type of local codes. LCCs can efficiently recover any coordinate of its corrupted encoding by probing only a few but not all fraction of the corrupted word. A q-ary LCC which encodes length k messages to length N codewords with relative distance Δ has three parameters: r, δ and ϵ. r is called query complexity recording the number of queries. δ is called tolerance fraction measuring the relative distance between encoding codewords and its corrupted codes which can be locally decoded. ϵ is called error probability showing the coordinate of its corrupted encoding fail to be recovered with probability at most ϵ. One fundamental problem in LCCs is to determine the trade-off among rate, distance and query complexity. But for a specific LCC, focus is on query complexity, tolerance fraction and error probability. Reed-Muller codes (RM codes) are the most presentative LCCs. In order to understand the "local" more clearly, we revisit local correctors for RM codes and analyze them in detail: 1)The decoding procedures; 2)The role of Reed-Solomon codes (RS codes) in decoding RM LCCs; 3)Other local correctors for RM codes. How parameters including r, δ and ϵ change in RM LCCs have been analyzed in different correctors. We believe this paper can help us understand local codes better and grasp the main soul of this research direction.