Challenges with Microcell Deployment & Configuration
Microcells and femtocells have been deployed by most major carriers as a way to allow their customers to deploy their own network(s) anywhere there is an Internet connection.
With cellular phones becoming not just voice, but now also data appliances, there is a growing need for operators to provide wide area (3-5 mile) coverage that assures high-rate data and voice services in nearly every location. What used to be considered “acceptable” gaps in coverage, like isolated homes, are no longer areas that carriers can overlook.
For this reason, microcells and femtocells have been deployed by most major carriers as a way to allow their customers, both businesses as well as individuals, to deploy their own network(s) anywhere there is an Internet connection. The sheer number of cells (large and small area), the inability for the carriers to control the position and use of them, and the handover between these ad hoc cells and the overall network create significant challenges in spectrum and interference management.
We use the term “microcells” to describe any cell smaller in size than the macrocell, which covers 2 km or more. In practice, most cells that cover less than 2 km are picocells and femtocells, whose coverage is below 200 m.
Macrocells that cover a number of miles almost always have some gap in the coverage between them. Yet, even if the entire area is covered with the signal of receivable strength, the throughputs at locations in-between cells are usually the smallest due to the lowest signal strength. Installing microcells in these areas resolve both of these issues. Special cases are the areas that are not conducive to the RF propagation, and therefore cannot be covered by a macrocell. This includes: tunnels, garages, underground railways, and large office buildings.
Macrocells can be overloaded by the amount of traffic they service, especially at peak use times or during special events (e.g. emergencies, concerts, etc.). One solution to is to add more capacity to the macrocell via additional assigned frequency channels. However, this solution has its limits, in that the service provider might not have any more licensed channels to use in this cell. If the significant source of traffic within the coverage of the macrocell is an isolated area (e.g., apartment building, office building, mall, train station), then installing microcells to serve this area is a simple way to increase the capacity.
A more general way to increase the capacity of the cellular system within the available spectrum is to create a layered/heterogeneous network. Such a network has a microcell underlay to allow for greater spectrum reuse that provides larger capacity and an overlay (umbrella) of macrocells that ensure continuity of the coverage between the microcells. The microcell layer would, in this case, have the majority of the service provider‘s licensed channels to allocate for its location, while the macrocell layer would have only a few channels to hold the users over until they move into an area covered by the microcell layer. To accomplish this type of deployment, the operator will set the handover parameters to prefer microcells over macrocells whenever a microcell is available and useable.
The main concern when deploying a microcell is availability of an appropriate location to install and backhaul the data from (connect it with the rest of the network). This is an even more significant problem with microcells, as they are deployed much more densely than the macrocells. It has become much more difficult to obtain a location for a macrocell due to local ordinances on the visual appearance of the antenna tower that a larger area cell site needs to operate.
Fortunately for microcell positioning, antenna towers are typically not required. The accompanying microcell electrical equipment is the size of a household appliance and operates off of ‘normal’ commercial/residential power grids. This significantly eases the microcell deployment.
Another low cost approach for deploying the microcells over a larger area is via remote radio heads. In such a system, the lower-level signal processing (baseband processing) is centralized. Baseband data is digitally distributed to the remote radio heads placed at the desired locations, where it is then converted to/from RF and transmitted/received via antennas. As they do not have the processing requirements of a normal cell site, remote radio heads are smaller size than a ‘typical’ microcell deployment. This is a good solution for tunnels, garages, underground railways, etc., where radio heads can be strung along a power/data wire.
Another important feature working in the microcell’s favor is willingness of the end users (e.g., those without cell coverage) to install microcells on their premises and provide backhaul via their broadband internet access. End users often voluntarily absorb all of the capital and maintenance cost of the cell site and, in many cases, this generates additional profit for the operator.
A macrocell needs to be configured for it to work with the rest of the cellular network and to not interfere with its ability to service other users. The primary concern is the selection of the provider-licensed frequency channels that it will use. This is a complex problem in the case of macrocells, and an even more challenging one for a network with a mix of macrocells and microcells. This mixed (heterogeneous) network is much more dynamic than the classical network structure that relies only on macrocells. For macrocell based networks, after initial deployment optimization, reconfiguration (retuning) is required only a few times a year.
Unlike classical cellular protocols (GSM, AMPS, and IS-136), the newer CDMA cellular networks reuse the same frequency channel among adjacent base stations. In this case, the power level of each base station broadcast signals and each end user signal is critical. In such systems, microcell power control needs to be tightly executed so they don’t interfere — not just with the macrocells, but also with adjacent microcells. To accomplish this, microcells must measure the power of the users connected to it and assess the power of the other cells operating on the same frequency channel.
More recent Orthogonal Frequency Division Multiple Access (OFDMA) cellular base stations, like Long Term Evolution (LTE) enhanced Node B’s (eNB’s), share parts of their frequency channel with multiple connected users. They coordinate with other adjacent base stations that use the same frequency channel and divide it up amongst the desired users.
In LTE, this is called Inter-Cell Interference Coordination (ICIC), under which the cell edge users are assigned parts of the frequency channel that are not used in the adjacent cells. An eNB can then transmit at a higher power in these parts of the frequency channel to serve the edge users properly without creating interference in the adjacent cells.
LTE microcells typically do not use some part of the frequency channel, which is reserved for the LTE macrocells. As the microcells do not service a large portion of the end users, this usually does not impact their throughput. However, it does provide macrocells with the part of the frequency channel, intended for their cell edge users, that is free from microcell interference.
This static setup is suboptimal, so LTE offers dynamic coordination between eNB’s to provide a better solution for the particular state of the network. Even though the ICIC communication between eNB’s is standardized, the eNB’s response to this communication is not defined. Therefore, dynamic LTE ICIC is especially challenging for a network where eNB’s are produced by different vendors. An LTE network with the mix of macrocells and microcells provides an additional challenge for dynamic coordination due to quantity and density of eNB’s that are often from different vendors.
Even the more basic aspects of cell configuration, like setting different cell parameters (e.g. Cell Identity, Paging Area codes, serving priorities for different classes of users, handover parameters, maximum output power, etc.) cannot be done manually, as was the case with the macrocells-only networks. In this case, it is not known where the microcell will be deployed before hand, and, during installation, this settings burden cannot be placed on the installer (often an operator’s end user customer).
Therefore service providers must adopt a Self Organizing Network (SON) concept of operation, where a centralized authority authenticates and configures the microcells the first time they are connected. By using SON, end users can operate and install microcells in a plug-and-play fashion with just some simple diagnostic lights on the box. SON also allows for a centralized microcell control during its operation, freeing end user from reconfiguring it.
If the end user installs the microcell of his selected cellular service provider on his or her own premises and broadband Internet connection, he or she may also control which cellular phones can access this microcell and with which priority. This is communicated to the cellular service provider who remotely configures the microcell to execute this without direct interaction of end user with the microcell.
In conclusion, microcells are a main way to increase the capacity of cellular networks. Their wide deployment is enabled by their size and ease of operation.