错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Towards Private Multi-operator Network Slicing

  • Blake Haydon,
  • Shangqi Lai,
  • Xingliang Yuan,
  • Sharif Abuadbba,
  • Carsten Rudolph

摘要

Network slicing allows customers to obtain networking resources down to the physical level and enables unprecedented resource isolation and fine-grained control. Such a technique is gaining adoption in current 5G networks and is poised to be influential in 6G and beyond. However, slicing may occupy finite spectrum resources, which require network operators to work cooperatively to optimise slice placement. Such cooperation raises critical privacy concerns. If mobile network operators share spectrum and customers, sensitive customer information can leak to cooperating operators throughout the slice optimisation process. To solve this issue, we construct three protocols of varying privacy levels for securely enabling network slicing among different mobile network operators. Our protocols employ efficient secure multiparty computation techniques to collaboratively optimise slicing without revealing customers’ identities or orders. We implement our protocols for the two-party setting using the MP-SPDZ framework. Our results highlight that for 32 slices on two basestations, we can produce allocations in under 5ms with a balanced security and efficiency requirement. Under the highest privacy level and similar parameters, our protocol generates a slicing allocation in under 7s. Similar to non-private methods, our solution finishes within the 60s heartbeat period of a Spectrum Access System, allowing for real-time updates with a changing spectrum allocation.