Quantum Machine Intelligence for 6G URLLC
Quantum Machine Intelligence for 6G URLLC
复制标题
DOI：
10.1109/mwc.003.2200382
复制
发表时间：
2023-04
影响因子：
12.9
通讯作者：
Fakhar Zaman;Ahmad Farooq;M. A. Ullah;Haejoon Jung;Hyundong Shin;M. Win
中科院分区：
文献类型：
--
作者：
Fakhar Zaman;Ahmad Farooq;M. A. Ullah;Haejoon Jung;Hyundong Shin;M. Win
虚拟现实或增强现实、触觉互联网、工业自动化和自主移动性等沉浸式和关键任务数据驱动应用正在为第六代（6G）网络中的超可靠和低延迟通信（URLLC）带来前所未有的挑战。机器智能采用深度学习、强化学习和联邦学习（FL），提供新的范例，以确保大数据训练流上的6G URLLC。然而，机器学习能力的传统局限性使得实现严格的6G URLLC要求具有挑战性。在本文中，我们利用量子资源的优势，例如叠加、纠缠和量子并行性，研究变分量子计算和量子机器学习（QML）在6G URLLC中的潜力。其基本思想是将量子机器智能与6G网络集成，以确保严格的6G URLLC要求。作为一个例子，我们展示了NP难URLLC任务卸载优化问题的量子近似优化算法。QML的变分量子计算也被用于无线网络中，以提高机器智能的学习速度，并确保关键任务应用的学习最优性。考虑到FL中的安全性和隐私性问题以及计算资源开销，进一步研究了量子辅助FL中的分布式盲量子计算和远程计算。
Immersive and mission-critical data-driven applications, such as virtual or augmented reality, tactile Internet, industrial automation, and autonomous mobility, are creating unprecedented challenges for ultra-reliable and low-latency communication (URLLC) in the sixth generation (6G) networks. Machine intelligence approaches deep learning, reinforcement learning, and federated learning (FL), to provide new paradigms to ensure 6G URLLC on the stream of big data training. However, classical limitations of machine learning capabilities make it challenging to achieve stringent 6G URLLC requirements. In this article, we investigate the potential of variational quantum computing and quantum machine learning (QML) for 6G URLLC by utilizing the advantage of quantum resources, such as superposition, entanglement, and quantum parallelism. The underlying idea is to integrate quantum machine intelligence with 6G networks to ensure stringent 6G URLLC requirements. As an example, we demonstrate the quantum approximate optimization algorithm for NP-hard URLLC task offloading optimization problems. The variational quantum computation for QML is also adopted in wireless networks to enhance the learning rate of machine intelligence and ensure the learning optimality for mission-critical applications. Considering the security and privacy issues, as well as computational-resource overheads in FL, distributed quantum computation in blind and remote fashions is further investigated for quantum-assisted FL.