ID 原文 译文
14375 最后，通过选取具有宏基站( Base Station，BS) 和小型 BS 的两层异构网络，对多智能体 DQN 算法的性能进行仿真实验。 Finally，the performance of the multi-agent DQN algorithm is simulated by selecting a two-layer heterogeneous network with macro base station( BS) and smallBS．
14376 仿真结果表明，相较于传统学习算法，多智能体 DQN 算法具有更好的收敛性，且能够有效提升用户设备( User Equipment，UE) 的服务质量与能效，并可获得最大的长期总体网络实用性。 The simulation results show that the multi-agent DQN algorithm has better convergence，and can effectively improve the service quality and energy efficiency of user equipment( UE) ． At the same time，the maximum long-term overall network utility can be obtained compared with traditional learning algorithms．
14377 针对视频监控系统接入层中前端设备的身份安全问题，通过对会话初始协议( Session Initiation Protocol，SIP) 进行研究和扩展，设计并改进了一种基于超文本传输协议( Hyper Text Transfer Protocol，HTTP) 摘要访问认证的 SIP 安全机制。 For the identity security of the front-end equipment in the access layer of video surveillance system，a session initiation protocol( SIP) security mechanism based on Hyper Text Transfer Protocol( HTTP)access digest authentication is designed and improved by studying and extending the SIP．
14378 前端设备在接入视频监控系统前，需要通过该安全机制与系统安全管理平台上的 SIP 服务器进行身份认证。 The front-end equipment needs to be authenticated with the SIP server on the system security management platformthrough a security mechanism before accessing the video surveillance system．
14379 认证双方基于公钥基础设施数字证书认证体系( Public Key Infrastructure /Certificate Authority，PKI/CA) 获取对方的数字证书后解析公钥，在摘要认证的基础上使用公钥加密和私钥签名来保护认证序列的安全性，解密认证序列后通过异或校验和摘要校验实现双向身份认证。 Based on the public key infrastructure digital certificate authentication system( PKI/CA) ，both parties obtain the digital certificate of theother party and parse the public key． According to the digest authentication，public key encryption and private key signature are combined to protect the security of the authentication sequence，and then the authentication sequence is decrypted and the bidirectional identity authentication is implemented through exclusive OＲ check and digest check．
14380 测试与分析结果表明，改进的安全机制能够抵御常见的 SIP 安全风险，实现设备与安管平台间的双向身份认证，在适当损失效率的情况下确保接入系统的设备身份合法可信。 The test and analysis results show that the improved mechanism can resistcommon SIP security risks and implement the bidirectional identity authentication between the equipmentand the management platform，ensuring that the identity of the equipment accessing the system is legal andcredible under the condition of proper loss of efficiency．
14381 现有音频增强现实( Audio Augmented Ｒeality，AAＲ) 服务类应用无法根据用户的交互行为进行动态调整音频引导内容，难以满足大规模人群多样化应用需求。 To solve the problem that the existing audio augmented reality( AAＲ) service applications cantdynamically adjust the audio guidance content according to the users interaction behavior，and cant meetthe diversified application needs of large-scale crowd.
14382 针对此问题，构建了一个基于AAＲ 的博物馆交互式引导模型，通过对场景、用户、导航过程三方面的定义实现音频信息的自适应推送，结合用户个人资料、行为特征、历史参观等信息匹配发现个人兴趣偏好，同时支持跟踪用户行为，动态调整兴趣特征参数，达到提供个性化的音频信息服务的目的。 This paper constructs an interactive guidance modelbased on AAＲ for museum． It realizes the adaptive push of audio information through the definition ofscene，user and navigation process，which can combine the users personal data，behavior characteristics andnavigation process． At the same time，it supports tracking user behavior and dynamically adjusting interestcharacteristic parameters to achieve the goal of providing personalized audio information service．
14383 实验结果表明，基于该模型的原型系统具有良好的实用性，能更容易地定位展品，并增加观众对展品的兴趣。 The experimental results show that the prototype system based on the proposed model has good practicability，can more easily locate the exhibits and increase tourists interest in the exhibits．
14384 针对现有驾驶员打电话行为检测方法存在精度低、实时性能差以及免提通话无法检测等问题,提出一种基于深度学习的多特征融合检测方法。 For the problems of low accuracy, poor real-time performance and inability to detect hands-free calls in the existing detection methods of drivers' calling behavior, a multi-feature fusion detection method based on deep learning is proposed.