Task description and Expected results
This task provides innovative methodologies for global energy optimization, such as dynamic decision making on sleep modes for radio network elements, on how to price services depending on the availability of renewable energy sources, and on how to manage interference between Small-cell Networks (SC Ns).
It will aim to develop the algorithms and protocols, and has been divided into two tasks. Task 10.1 covers the two right-hand side boxes of Fig. 1 while Task 10.2 is focused on looking at the optimization problem from an upper, i.e., a global level perspective (integration). The algorithms for energy efficiency and the global optimization aspects is milestone M6.
Task 10.1 C haracterization of innovative optimization features in Small Size Green C ell Networks
As mentioned earlier, one of the major aims of this project is to advance the use of energy efficient SC Ns. This task identifies and characterizes the main optimization features to be considered in a joint energy optimization scheme:
1) C ell zooming and relaying
2) Dynamic sleep modes and their impacts – network level policies might decide to switch off certain macro cells, whereby a central entity such as cell zooming server could make these types of decisions. By zooming out the cells during low traffic demand, some equipment in the cell such as relays and pico-cells can be switched off while coverage is still guaranteed by the macro BS.
3) Interference management between macro-cells and the SC s, and also among the SC BSs – having information about cell positions and transmission powers, interference management can be based on a specific model that describes the uplink/downlink interference between macro BSs and SC s, as well as among small-cell BSs.
Task 10.2 Global Energy C onsumption Optimization
In this task, we look at the energy consumption optimization problem from an upper level perspective, i.e., a global perspective, while also keeping in mind renewable energy aspects and parameters.
In green communications, energy management is a multidimensional optimization problem, which consists of dynamically controlling a set of parameters to minimize the average energy consumption under several performance constraints. These constraints are associated with the user traffic demand (data rates and QoS), available transmission technologies, distance and attenuation conditions to the available base stations or femto-cells, cost of the spectrum bands, network features such as the use
of cooperation and the type energy used to supply the network/radio equipment.
The solution of the energy consumption minimization problem is an effective allocation of the available physical resources and a dynamic definition of the optimization parameters (frequency band, cell zooming/interference, active/sleep cells, e.g., by BS wilting). This demands for a global real-time scheduler that selects the best set of available resources.
The optimization scheme will be based on the following aspects:
1) User demand (schedule the resources required by the different users)
2) Energy Pricing (the different types of energy used to supply the plethora of radio and network equipment)
3) Optimization Methodology (analysis of the available radio spectrum opportunities and its energy cost, prioritization of the scheduling of more unoccupied radio bands that are simultaneously supplied by green-energy; definition of the transmitting power of each cell for a specific band by considering the users requirements, the interference management constraints, the energy prioritization/pricing and the specific network topology).
Switching off network equipment is particularly important if the source of energy has a high carbon (C O2) footprint. However, if the power is provided by renewable sources, the network elements may remain operational. Therefore, the decision to switch off certain network elements should also take the source of energy of the associated equipment into account.
Members of the research team in this task: (BI) Bolseiro de Investigação (Mestre) 3; (BPD) Bolseiro de Pós-Doutoramento 1; António Jorge da Silva Morgado; Daniel Luís
Silveira Robalo; Fernando José da Silva Velez; Norberto José Gil Barroca;
