Leveraging geo-located data and automation for improved network planning
Since the advent of mobile networks, network planning has been the fundamental first step upon which a high performing network was based. While tools and technologies have evolved and methodologies have been tweaked, the process has remained largely the same for over 20 years. In brief, network planning relied on utilizing detailed terrain, clutter and population maps coupled with marketing predictions, to design a network with the required population coverage and capacity to deliver on the company’s business plan.
The traditional planning process was designed to primarily deploy greenfield (new coverage) macro sites. In recent years as most Communication Service Providers (CSP’s) achieved a level of network maturity, the focus of planning has shifted from network roll-out to co-locating new technologies adding macro capacity. Focus is now shifting back to new site deployment, but in the form of small cells. This brings a couple of challenges for which the traditional planning process and tools are not well suited. Cells now cover 10’s of meters, not kilometers, and the number of small cells required is exponentially more than was the case with macro cells.
Other factors are also driving a change in the requirements of network planning engineers: As customers have grown to rely more on their mobile devices, their expectation of network performance has increased and coverage holes and quality issues are no longer tolerated. To deliver on these increased expectations requires additional sites to be built. In opposition to this though, increased market competition means the budget available for network expansions is under severe pressure. Planning engineers therefore need, more than ever, to determine the most profitable and efficient locations to deploy new sites ensuring maximum return on investment for each and every site.
Leveraging geo-located data
For a number of years geo-located usage data has been used with great effect to optimize networks when the lack of location accuracy and subscriber level granularity from OSS performance management data began to limit optimization effectiveness. The need to plan ever smaller areas, due to the densification of networks, means that just “cell-wide” is no longer accurate enough, but leveraging geo-located data for increased granularity provides an excellent solution.
To take advantage of this, TEOCO created ASSET Geo, a module for our ASSET radio planning tool. ASSET Geo enables planning engineers to visualize geo-located traffic and performance data within ASSET, increasing the precision with which they can place new sites for capacity offload or to solve network performance issues, ensuring every new site delivers maximum return on investment.
ASSET Geo is technology agnostic and supports CDMA, GSM, UMTS and LTE. It leverages the geo-location capabilities of TEOCO’s Geo Server which includes advanced positioning algorithms, and the ability to isolate stationary, moving, indoor and outdoor traffic allows for the creation of highly accurate traffic maps. These traffic maps consist of multiple layers and show traffic on a per-bearer, per-service or per-UE basis. Leveraging these maps allows not only highly accurate placement of new sites but also analysis of specific services or bearers, allowing a planning engineer to design for specific usage and not just traffic volume.
Adding additional data sources
While geo-located data provides excellent insights into the usage and location of your subscribers, leveraging social media and crowdsourced data can provide even further insights. The question that cannot be answered with geo-located data alone is, in areas where there is little to no traffic, is this because it is a low traffic area or do you perhaps have unidentified coverage issues; or are your competitors particularly strong in that area due to a number of large corporate accounts or a strong marketing push?
By using geo-located social media maps (from Twitter, for example) or crowdsourced data (such as from OpenSignal) which highlights subscriber activity from all networks and then overlaying this data on geo-located traffic maps from your network, it is possible to identify areas where you have low traffic, yet traffic hotspots still exist. These discrepancies could be the result of poor coverage on your network or a strong presence from a competitor and can provide valuable insight into areas which may require further investigation and focus.
Automating small cell planning
Does deploying a small cell make more sense than adding capacity to a macro cell? With a strong geo-located network usage dataset, that question is easier to answer. A more difficult question though is, where should I deploy x number of small cells for maximum impact, where maximum impact consists of capacity, coverage and VIP customer focus factors derived from a business strategy?
With many CSP’s turning to small cells to address the capacity demands of their customers, this is becoming a much more common question. Planning hundreds of small cells is a huge task, but with an excellent traffic dataset and well-defined performance and business criteria on which to base the plan, automation is an obvious choice.
To achieve this, TEOCO created Velox. Velox provides a single purpose built solution for small cell planning, optimized for speed and the evaluation of thousands of small cell candidates concurrently. Velox leverages geo-located traffic maps to determine an optimal radio design. Backhaul is, however, a key cost and performance consideration in a small cell network. Uniquely, Velox is also able to determine the backhaul requirements for small cell candidate locations. By assessing both the performance of the radio design and the cost of the backhaul for each candidate location during the design optimization process, Velox is able to determine optimal design, balancing cost and performance.
In conclusion
By using ASSET Geo to visualize geo-located traffic and performance data in ASSET, and Velox to automate the radio and backhaul designs for small cell networks, engineers can dramatically improve the accuracy and efficiency of their network and small cell planning.