Cloud computing is now used by many enterprises due to its increased computational efficiency, economic effectiveness, as well as flexibility. However, security is currently the main issue impeding the cloud computing platform's growth. Therefore, Decentralized Information Flow Control (DIFC) has been proposed as a suitable remedy for resolving the cloud security problems. Using conventional network access and encryption technology was not practicable in the DIFC to effectively restrict the spread of the tenant's personal data inside the system. Therefore, a novel DIFC framework for cloud virtual machines (VM) is suggested here. The suggested system encapsulates four entities such as central authority (CA), encryption proxy (EP), cloud server (CS), and cloud tenant VM. The EP has implemented the ciphertext data-flow security technique. Encryption is carried out using the newly proposed hybrid “Advanced Encryption Standard (AES)–Elliptic Curve Cryptography (ECC) algorithm”. The hybrid AES-ECC encryption technique uses the proposed Improved Poor Rich Optimization (IPRO) model to compute the optimal key. The implementation of the developed work is evaluated against the existing works for the "Chess, T1014D100K, and Retail datasets”. In particular, for the T1014D100K dataset, the cost function of the suggested model at the 2.5th iteration is 57.14%, 62.05%, 80%, 54.2%, and 56% better than the old models like BOA, SMO, SSA, PRO, and LA correspondingly.

云计算由于其提高的计算效率、经济效益和灵活性，现在被许多企业使用。然而，安全是目前阻碍云计算平台发展的主要问题。因此，分散信息流控制（DIFC）被提议作为解决云安全问题的合适补救措施。使用传统的网络访问和加密技术在 DIFC 中无法有效地限制租户个人数据在系统内的传播。因此，这里建议了一种用于云虚拟机 (VM) 的新型 DIFC 框架。建议的系统封装了四个实体，例如中央机构（CA）、加密代理（EP）、云服务器（CS）和云租户VM。EP 实现了密文数据流安全技术。使用新提出的混合“高级加密标准（AES）-椭圆曲线密码术（ECC）算法”进行加密。混合 AES-ECC 加密技术使用建议的改进穷富优化 (IPRO) 模型来计算最佳密钥。针对“国际象棋、T1014D100K 和零售数据集”的现有工作评估开发工作的实施。特别是，对于 T1014D100K 数据集，建议模型在第 2.5 次迭代时的成本函数为 57.14%、62.05% , 80%, 54.2%, 和 56% 分别优于 BOA, SMO, SSA, PRO, LA 等老型号。混合 AES-ECC 加密技术使用建议的改进穷富优化 (IPRO) 模型来计算最佳密钥。针对“国际象棋、T1014D100K 和零售数据集”的现有工作评估开发工作的实施。特别是，对于 T1014D100K 数据集，建议模型在第 2.5 次迭代时的成本函数为 57.14%、62.05% , 80%, 54.2%, 和 56% 分别优于 BOA, SMO, SSA, PRO, LA 等老型号。混合 AES-ECC 加密技术使用建议的改进穷富优化 (IPRO) 模型来计算最佳密钥。针对“国际象棋、T1014D100K 和零售数据集”的现有工作评估开发工作的实施。特别是，对于 T1014D100K 数据集，建议模型在第 2.5 次迭代时的成本函数为 57.14%、62.05% , 80%, 54.2%, 和 56% 分别优于 BOA, SMO, SSA, PRO, LA 等老型号。