Journal of Frontiers of Computer Science and Technology ›› 2023, Vol. 17 ›› Issue (4): 973-984.DOI: 10.3778/j.issn.1673-9418.2107053

• Network·Security • Previous Articles    

Hesitant Fuzzy Method of Rewarding Good and Penalizing Bad in Cloud Service User Behavior-Based Safety Evaluation

PENG Dinghong, SONG Bo, ZHANG Wenhua   

  1. Institute of Quality Development, Faculty of Management and Economics, Kunming University of Science and Technology, Kunming 650093, China
  • Online:2023-04-01 Published:2023-04-01

云用户行为安全评价的犹豫模糊奖优罚劣方法

彭定洪,宋博,张文华   

  1. 昆明理工大学 管理与经济学院 质量发展研究院,昆明 650093

Abstract: Security issue is one of key contents of the related cloud computing research, and unsafe behavior of cloud computing service users (CSU) is the main cause of cloud computing security threats. The behavior-based safety (BBS) evaluation of CSU is the basis for correcting unsafe behaviors, maintaining safe behaviors, and effectively improving cloud computing security. In order to effectively evaluate behavior-based safety (BBS) of CSU and reflect the purpose of rewarding good and penalizing bad in evaluation process, a hesitant fuzzy CSU-BBS evaluation method based on the idea of TOPSIS (technique for order preference by similarity to an ideal solution) virtual worst solution is proposed. Firstly, differential CSU behavior data obtained by multi-point monitoring are expressed in hesitating fuzzy element (HFE) and directly used for subsequent evaluation to ensure that the CSU behavior data used for evaluation are true and comprehensive. Secondly, in view of the characteristics of “reward and penalization distinct” of security issues and the need of “positive and negative reinforcement” of BBS, a hesitant fuzzy de-dimensional method that reflects the function of “rewarding good and penalizing bad” is proposed. Next, in order to obtain a reasonable ordering of multiple schemes, under the virtual worst solution (TOPSIS variant) framework, a hesitant fuzzy procedure for evaluating CSU-BBS is developed. Finally, the BBS of five CSU is evaluated based on the user behavior data of three cloud computing platforms of a small and medium-sized Internet company. Evaluation results show that the proposed method can effectively evaluate CSU-BBS and has the function of rewarding good and penalizing bad.

Key words: cloud computing, behavior-based safety (BBS) evaluation, hesitant fuzzy set (HFS), “rewarding good and penalizing bad” de-dimensional method, virtual worst solution

摘要: 安全问题是云计算相关研究中的关键议题,云计算用户(CSU)的不安全行为是云计算安全威胁的主要致因。对CSU进行行为安全(BBS)评价是纠正不安全行为、维持安全行为,从而有效提升云计算安全的基础。为有效对CSU-BBS评价且体现安全评价的奖优罚劣目的,提出一种基于逼近理想解法(TOPSIS)虚拟最劣解思想的犹豫模糊奖优罚劣的CSU-BBS评价方法。首先,对于多点监测获取的差异化CSU行为数据以犹豫模糊元(HFE)形式表达,并直接用于后续评价,确保用于评价的CSU行为数据真实、完全;其次,针对安全问题应“奖罚分明”的特点及BBS的“正负强化”需要,在犹豫模糊背景下提出一种体现“奖优罚劣”功能的犹豫模糊去量纲方法。接着,为获得多方案合理排序,基于TOPSIS的变体——虚拟最劣解方法,发展了一种CSU-BBS优劣评价的犹豫模糊求解途径。最后,以某中小型互联网企业的3个云计算平台用户行为数据对5名CSU进行BBS实例评价并辅以对比分析。评价结果表明所提方法能够有效评价CSU-BBS且具奖优罚劣功能。

关键词: 云计算, 行为安全(BBS)评价, 犹豫模糊集(HFS), “奖优罚劣”去量纲, 虚拟最劣解