Federated Learning ensures that clients can collaboratively train a global model by uploading local gradients, keeping data locally, and preserving the security of sensitive data. However, studies have shown that attackers can infer local data from gradients, raising the urgent need for gradient protection. The differential privacy technique protects local gradients by adding noise. This paper proposes a federated privacy-enhancing algorithm that combines local differential privacy, parameter sparsification, and weighted aggregation for cross-silo setting. Firstly, our method introduces Renyi differential privacy by ´ adding noise before uploading local parameters, achieving local differential privacy. Moreover, we dynamically adjust the privacy budget to control the amount of noise added, balancing privacy and accuracy. Secondly, considering the diversity of clients’ communication abilities, we propose a novel Top-K method with dynamically adjusted parameter upload rates to effectively reduce and properly allocate communication costs. Finally, based on the data volume, trustworthiness, and upload rates of participants, we employ a weighted aggregation method, which enhance the robustness of the privacy framework. Through experiments, we validate the effective trade-off among privacy, accuracy, communication costs and robustness achieved by the proposed method.

Abstract Due to the limitations of physical resources, if a large-scale emulation network environment composed of millions of vertices and edges is constructed by virtualization technology, the whole network topology should be partitioned into a set of subnets. The topology partition is a work of graph partition. The existing topology partition methods have shortcomings, such as low efficiency and poor practicability, especially for large-scale network topology. The emulation network is a kind of complex network and has the characteristics of community structure. Therefore, we proposed LENTP (large-scale emulation network topology partition) based on the community detection with the weight of the vertex similarity for large-scale topology partition. In the first stage, the tree-structured area compression reduces the topology scales significantly to improve partition efficiency. And then, the improved Louvain algorithm is used to topology partitioning and obtain an initial set of subnets with the minimum number of subnets and remote links. Finally, after repartitioning and merging for the initial subnets, the result of subnets is the final topology partition that reaches the optimization objectives with the conditions of the virtual resources. In the experiment, the method is tested in five groups of network topology with different scales. The results demonstrate that LENTP can partition the network topology over 1 000 000 nodes and significantly improve the running-time efficiency of the network topology partition.

Relational databases are widely applied in various industries, such as government departments, medical institutions and enterprises, for data storage and relationship management. It is convenience to data maintenance, however, the data in it maybe vulnerable to be forged or tampered with. Therefore, protecting the copyright of relational databases is a critical issue. Fortunately, watermarking technology can be used to prove copyright ownership. Since the digital watermarking technology can solve this problem effectively, based on semantic, we propose two watermarking approaches with high security and strong robustness for numeric and non-numeric data of relational databases in this paper. On the one hand, we propose a reversible numeric watermarking approach. It performs the replacement operation at the semantic level and retain the statistical characteristics of the data. On the other hand, based on word segmentation and word embedding, the non-numerical attributes of the relational database with natural language are chosen to embed watermark. It is noteworthy that the proposed mechanism can be applied to both Chinese and English with minimum distortion. Additionally, we also propose the virtual splitting of attribute column and the principle of modification minimum to guarantee the capacity of watermark and reduce the data modification rate. Additionally, the BCH (31,16,7) error control code is added to the binary watermark string to improve the detection rate of watermark. Based on the above innovations, the security of watermarking algorithm is improved successfully by double encryption (chaos encryption and hash encryption) with two keys (the user's private key and the attribute column key). The simulation results demonstrate that the proposed two algorithms in this paper have stronger robustness on defensing malicious attacks than previous works.

Abstract As the backbone of the domain name system (DNS), DNS resolvers are essential to the Internet. Nowadays, the measurement of DNS resolvers, especially open DNS resolvers, has become a research focus. Previous research works show that DNS responses returned from some open DNS resolvers are not expected for clients and the Internet. We call these DNS responses ‘unexpected DNS responses’. Research on unexpected DNS responses is beneficial to the research, usage and management of open DNS resolvers. This paper explores unexpected DNS responses returned from open DNS resolvers in terms of identification and classification to better understand the behaviours of open DNS resolvers. First, an identification method is proposed to identify all kinds of DNS responses from each section of DNS messages. Second, a classification method is proposed to classify unexpected DNS responses by their influences on clients and the Internet. Furthermore, an efficient identification and classification method is proposed to simplify the above process. Among about 9 million responding open DNS resolvers in the experiments on the IPv4 address space, about 40% return unexpected DNS responses. Experimental results show that the proposed methods can identify and classify all kinds of DNS responses returned from open DNS resolvers.

The Domain name system (DNS) is crucial to Internet services. Previous studies have pointed out that the name dependency in domain name resolution poses a risk to the security of DNS. In this paper, we present measurement results from a dataset containing resolution paths of domain names collected from a large-scale survey. This dataset is used to research the effect of the name dependency on the DNS, reaffirm findings in published work, and notice some significant differences. When name resolution spans multiple domains, it will lead to name dependency and make the resolution process more complex. Furthermore, we assess the risk of domain name resolution: a name resolution fault analysis model and the calculation of the failure probability of name resolution is proposed. The model can identify the key server sets that lead to the resolution failure of a domain name, and quantify the failure probability of its resolution. This research provides a breakthrough point for guiding the configuration, management, deployment, and upgrading of the DNS.

With the development of big data, relational databases are playing an important role in enterprises, military affairs, medical, etc. Moreover, they are vulnerable to piracy, forgery, and tampering. Consequently, the copyright protection of relational databases has become an issue with increasing concern. The use of digital watermarking technology can solve this problem effectively. In this paper, by using Chinese word segmentation and word embedding, the non-numerical attributes of the relational database with Chinese natural language are chosen to embed watermark. The method of the virtual splitting of the attribute column and the principle of modification minimum are proposed to guarantee the watermark capacity and reduce the data modification rate. The simulations are carried out to prove the algorithm has strong robustness to defense malicious attacks.

Mobile edge computing (MEC) has emerged for reducing energy consumption and latency by allowing mobile users to offload computationally intensive tasks to the MEC server. Due to the spectrum reuse in the network of MEC, the inner-cell interference has a great effect on MEC's performance. In this paper, for reducing the energy consumption and latency of MEC, we propose a game theory based approach to join task offloading decision and resource allocation together in the MEC system. In this algorithm, the offloading decision, the CPU capacity adjustment, the transmission power control, and the network interference management of mobile users are regarded as a game. In this game, based on the best response strategy, each mobile user makes their own utility maximum rather than the utility of the whole system. We prove that this game is an exact potential game and the Nash equilibrium (NE) of this game exists. We also investigate the properties of this algorithm, including the convergence, the computational complexity, and the Price of anarchy (PoA). We evaluate the performance of this algorithm by simulation. The simulation results illustrate that this algorithm is effective in improving the performance of the multi-user MEC system.

In the process of resolving domain names to IP addresses, there exist complex dependence relationships between domains and name servers. This paper studies the impact of the resolution dependence on the DNS through constructing a domain name resolution network based on large-scale actual data. The core nodes of the resolution network are mined from different perspectives by means of four methods. Then, both core attacks and random attacks on the network are simulated for further vulnerability analysis. The experimental results show that when the top 1% of the core nodes in the network are attacked, 46.19% of the domain names become unresolved, and the load of the residual network increases by nearly 195%, while only 0.01% of domain names fail to be resolved and the load increases with 18% in the same attack scale of the random mode. For these key nodes, we need to take effective security measures to prevent them from being attacked. The simulation experiment also proves that the resolution network is a scale-free network, which exhibits robustness against random failure and vulnerability against intentional attacks. These findings provide new references for the configuration of the DNS.