The provision of very high system capacity (traffic per square meter) and very high per user data rates requires a densification of the radio-access network (As earlier stated in THIS ARTICLE) by deploying heterogeneous networks. The introduction of these networks causes additional interference problems.
In itself, the use of heterogeneous network deployments in mobile-communication systems – that is, complementing a macro layer with lower-power pico nodes to increase traffic capacity and/or achievable data rates in specific areas – is nothing new and has been used for a relatively long time in, for example, GSM networks. Different sets of carrier frequencies have then typically been used in the different cell layers, thereby avoiding strong interference between the layers. However, for a wideband radio-access technology such as LTE, using different carrier frequencies for different cell layers may lead to an undesirable spectrum fragmentation.
The simultaneous use of the same spectrum in different cell layers will obviously imply interference between the layers. Due to the difference in transmit power between the nodes of a heterogeneous network deployment, such inter-layer interference may be more severe compared to inter-cell interference between cells of the same layer.
The characteristics of the inter-layer interference in a heterogeneous network deployment will also depend on the exact cell-selection strategy being used. Conventional cell selection is typically based on terminal measurements of the received power of some downlink signal, more specifically the cell specific reference signals in the case of LTE. However, in a heterogeneous network deployment with cells with substantially different transmit power, including different power of the reference signals, selecting the cell that is received with the highest power implies that the terminal may often select the higher-power macro cell even if the path loss to a pico cell is significantly smaller. This will obviously not be optimal from an uplink coverage and capacity point of view. It should also be noted that, even in terms of downlink system efficiency, it may not be optimal to select the cell with the highest received power in a heterogeneous network deployment. Although the high-power macro cell is received with higher power, this is at least partly due to the higher macrocell transmit power. In that case, transmission from the macro cell is associated with a higher “cost” in terms of interference to other cells. Expressed alternatively, a transmission from the macro cell will prohibit the use of the same physical resource in any of the underlaid pico-cells. Alternatively, at the other extreme, cell selection could be based on estimates of the (uplink) path loss. In practice this can be achieved by applying a cell-specific offset to the received power measurements used in conventional cell selection, an offset that would compensate for the difference in cell transmit power. Such a cell-selection strategy would extend the area in which the pico cell is selected, as illustrated in the illustration below.
It is therefore also sometimes referred to as range extension. Selecting the cell to which the path loss is the smallest – that is, applying range extension – would maximize the uplink received power/SINR, thus maximizing the achievable uplink data rates. Alternatively, for a given target received power, the terminal transmit power, and thus the interference to other cells, would be reduced, leading to higher overall uplink system efficiency. Also, it could allow for the same downlink physical resource to also be used in other pico cells, thereby also improving downlink system efficiency. However, due to the difference in transmit power between the cells of the different cell layers, there is an area where the pico cell is selected while, at the same time, the downlink transmission from the macro cell is received with substantially higher power than the actual desired downlink transmission from the pico cell. Within this area, there is thus potential for severe downlink inter-cell interference from the macro cell to pico-cell terminals, interference that may require special means to handle.
One possible approach to handle the extended interference in a heterogeneous network deployment with range-extended pico cells is to use carrier aggregation in combination with cross carrier scheduling. The basic principle of such an approach is illustrated in the illustration for the case of two layers (macro and pico) and two downlink carriers (fM and fP).
If the carrier-aggregation approach described above cannot be used, one can instead apply more conventional interference coordination between the different layers, similar to interference coordination within one cell layer, but extended to also cover interference between control-channel transmissions of the different cell layers.
With such an approach, the same carrier is used for transmission in both the macro layer and the under-laid picocells. However, the power of macro-cell transmissions is restricted in some subframes. This is illustrated below;
