Urban small cells, also referred to as outdoor picocells, microcells and metrocells, are a key tool for oper­ators to increase capacity and coverage depth in areas of high traffic demand density. These are operator deployed and managed, and enhance mobile broadband connectivity and quality of experience (QoE) in public spaces, both outdoors and indoors. However, providing cost-effective backhaul is considered a key challenge.

A look at operators’ motivations as well as perceived barriers to the roll-out of urban small cells…

Developing backhaul requirements for urban small cells

For small cell backhaul, there is no one-size-fits-all solution and often a trade-off between capacity, coverage and cost needs to be made. There are various factors that impact backhaul deployments. These include hotspot capacity, wide area capacity for enhanced user QoE, indoor coverage depth, access technologies such as 3G, long term evolution (LTE), Wi-Fi or their combinations thereof, access spectrum (only the total bandwidth impacts backhaul capacity requirements; little sensitivity to the radio access network (RAN) carrier frequency is noted if the frequency ranges currently utilised for mobile RANs are considered), and HetNet coordination (dedicated or shared carrier for small cells).

  • Impact of deployment motivation on backhaul requirements: From a backhaul perspective, relaxations can be made depending on the use case. For a capacity-driven small cell deployment, backhaul capacity should not limit the throughput of the small cell. In the capacity case, existing macro coverage is assumed and so there is potential for relaxed availability for small cells. Over­lapping macro-small cell coverage may require coordination for handover and resource usage, which drives tighter synchronisation and, in turn, tighter backhaul delay performance.
  • Small cell backhaul capacity provisioning: Backhaul capacity is a key design decision and, when outsourced, an important part of the service level agreement (SLA). Different approaches are needed for coverage-driven versus capacity-driven deployments. In a coverage-limited scenario typical of enterprise small cells, the latter do not run at full capacity and so provisioning can be determined by the end-user traffic demands. In a capacity-limited case, it is assumed that demand exceeds the cap­ability of small cells and the backhaul capacity can be determined by the limitations of these cells. In cases where an operator has limited spectrum to increase macro capacity, urban small cells are likely to be deployed. Hence, sharing of the RAN spectrum between small cells and macro is likely.
  • Transport security: The traffic carried over small cell transport should be protected against unauthorised intrusion and tampering. Some types of small cells, including 3G and Home-eNodeB classes, have mandatory encryption on their backhaul interfaces and thus remain protected by default. For the eNodeB class of small cells, the 3rd Generation Part­ner­ship Project states that transport encryption is required only for those transport segments not considered “trus­­t­ed” by the operator. Transport which is inherently secure can therefore avoid the bandwidth overhead of IPsec. Urban small cell networks may comprise segments outsourced to service providers, such as fibre or cable owners, which might not be considered trusted from an operator’s perspective. IPsec encryption terminates in the small cell at one end and in a security gateway at the other, which may reside in the small cell core network shown in the reference network architecture.
  • Backhaul requirements to support co­ordinated HetNets: Time domain co­ordination between small cells and macro cells is needed for time division duplex networks and for HetNet coordination. This requires phase synchronisation between coordinated cells and signalling between cells over the backhaul to set up and adjust the coordination parameters. Of these, it is the phase synchronisation that may drive stringent backhaul delay performance requirements where packet synchronisation techniques are used (as opposed to global positioning system).

Commercial models

Urban business models can take several forms and encompass different use cases, from operator installed, owned and maintained, to variants where customers manage part of the activities, and models where a third party provides all activities on behalf of multiple operators and customers.

An operator self-deployed approach is the one where an operator plans, designs, builds and operates the network. In this case, the operator is responsible for site acquisition, permitting, installation, integration, backhaul provisioning, optimisation and operations.

Operator site share, on the other hand, is characterised by multiple operators working together to share common infrastructure. A possible scenario in this case is that one operator will obtain access to a vertical structure for the installation of the small cell or small cell antennas and share access with additional operators. Typically, each operator will provide their own installation, integration, backhaul provisioning, optimisation and operations support. The level of technological complex­ity will dictate the degree to which assets can be shared.

Third-party neutral host encompasses a single organisation providing services for a number of operators wishing to install infrastructure at that location. In this model, the third-party neutral host interfaces with multiple operators to provide a solution that is amenable to the operators as well as to the local authorities. A common scenario for the third-party neutral host is to provide access to the vertical asset, permitting, installation, power supply and, in some cases, backhaul connection. The operator or small cell hardware vendor is responsible for integration, optimisation and monitoring of these components into the macro network.

The third-party-leased (wholesale) capacity business model is similar to an extensive neutral host proposition. Here, all activities such as designing/planning, building and operating are undertaken by the third party.

Backhaul selection, planning and deployment

Small cell backhaul (SCB) refers to the transport between the small cell and a point of presence (PoP), which has a connection with the operator’s core network. In general, the SCB comprises a mix of wireline connectivity and a number of wireless technologies. As there is no defin­ite SCB solution that suits all conditions, operators will need to use a toolkit. The following steps help in selecting a toolkit and planning guidelines to decide which tool to use.

  • Guidelines should be developed to simplify the subsequent planning of indivi­dual networks. These should include a range of conditions in the markets in which an operator deploys backhaul technology options and factors such as budget, lifespan and zoning restrictions. A wide range of possibilities is reduced to a toolkit of backhaul solutions with rules to select the appropriate tool for a given set of conditions. Examples could include minimum and maximum limits on the number of backhaul extensions attached to a PoP to ensure sufficient amortisation costs without limiting the small cell’s performance; plugging in operator and region-specific total cost of ownership data for each backhaul option; and the development of flow diagrams to facilitate decision-making.
  • Planning applies the guidelines to a specific city or market in order to identify the set of small cell sites needed to meet coverage and capacity requirements, and the type of backhaul to be used for each. The market-specific input data needed for planning will include a spatial forecast of unserved demand, and a list of candidate sites and locations of PoPs. Planning will involve using computer-aided design tools to draft different components of the network. Adherence to the guidelines should result in the desired network improvement being achieved within the TCO budget. The plan may incorporate a phased deployment in line with the demand growth forecast. The refinement or waiver of certain rules is likely to be needed during early planning and for special situations.
  • Deployment is the process of installing and commissioning at chosen sites according to the plan. Acceptance testing revealing the actual achieved network and backhaul performance may point to shortfalls to be addressed through subsequent updates in the plan.
  • Operation maintains the network to deliver target service levels with ongoing maintenance and planned upgrades. This is a continuous cycle rather than a one-shot process, with updates needed to refine processes and keep up with the changes in market conditions. However, it is also desirable to develop guidelines and plans with a reasonable shelf life so that teams involved with the subsequent stages are able to benefit from a period of stability.

Synchronisation for urban small cells

In the urban environment, additional functionalities such as enhanced inter-cell interference coordination (eICIC) may be required to manage interference between macro-eNodeBs and pico-eNodeBs. eICIC, even for frequency division duplexing radio technologies, requires a 1-5 user phase synchronisation between the slaves on the macro-eNodeBs and the pico-eNodeBs.

For indoors, the global navigation satellite system signal may be too weak to penetrate buildings. Therefore, in such cases, a packet-based synchronisation solution such as PTP is one of the most practical and generic options available.

One of the key challenges for urban small cells is the choice of last mile backhaul. This choice is also of paramount importance for the synchronisation performance that can be achieved. It is important to take into account challenges presented by different network access technologies in the choice of backhaul transport technology.

Based on a white paper, “Backhaul for Urban Small Cells”, released in June 2015 by the Small Cell Forum, supported by the Metro Ethernet Forum