Dynamic resource allocation and pricing for shared radio access infrastructure

Service-Aware Network Slice Trading in a Shared Multi-Tenant Infrastructure

“…The findings in [10] also showed that the size of the time window plays a significant role in the ability of the framework to adapt to network fluctuations. Using a metric called the "Relative Distance to Optimum" (RDO), [10] showed that the best value was = 100 ms.…”

“…Namely, this constraint ensures that a given tenant's gap is the difference between the tenant's utility target (i.e., th, ) and the achieved utility. Though visually similar to the formulation in [10], note that a tenant's achieved utility -in this formulation -is calculated as the sum of the utilities of all the tenant's services catered to. 3 The individual service utilities are computed using the utility functions illustrated in Fig.…”

“…Furthermore, we assume that each tenant pays a cost per assigned resource, which is composed of three parts, namely: capital expenditure (CapEx) represented by ca ; operational expenditure (OpEx) denoted as op ; and finally, the pressure cost given by pre . As discussed in [10], the pressure cost links the tenants' gaps (i.e., [ ]) with the revenue necessary for expansion.…”

“…The total number of slots contained in the entire time period of operation (during which the optimization is to be performed) is denoted by . For the sake of clarity and continuity, this work coopts the notations as used in [10]. Namely, the fraction of resources assigned to a user at time slot is represented by [ ].…”

“…As a result, operators in such a framework can often find themselves in situations of resource surplus, where they incur unnecessary expenditure by paying for unused resources, or resource scarcity, where they risk having dissatisfied customers. To address this issue, our previous work [10], while still relying on state-of-the-art SLAs and considering active sharing, provides a techno-economic model that permits short-term dynamic resource trading (i.e., on the order of seconds/minutes), wherein the mobile virtual network operators (MVNOs) can buy or sell resources based on their customers' needs and, as a consequence, deviate from the original SLA to a certain extent. While the idea proposed in [10] works quite well when the MVNOs happen to choose similar types of services, it struggles to accommodate scenarios wherein the service heterogeneity is large.…”