ISCO Value Model Quantifies Revenue from Spectrum Conditioning

The Situation

Up to now the value of ISCO spectrum conditioning solutions has been derived from the intrinsic value of improving the subscriber experience by reducing dropped calls, reducing ineffective attempts, improving data rates, reducing RSSI or RTWP and many others. All very positive benefits but they are hard to quantify in dollars therefore difficult to justify deployment beyond an "as needed" spot basis. Our customers requested help justifying a wider deployment.

The Resolution

ISCO has developed a model to help our customers financially justify a wider deployment of spectrum conditioning across their networks. Based on the numerous devices deployed and subsequent performance analysis conducted across the installed base, ISCO has created the ISCO Value Model (IVM).

The core premise is based on the growing demand to optimize wireless networks and increase capacity through improved spectrum utilization. The ISCO Value Model gives the wireless operator a hard dollar value that will be realized from the deployment of ISCO's Proteus spectrum conditioning solution on a wider scale.

To begin, the IVM makes statistical assumptions about the geographic and temporal distribution of interference in the network. From a network perspective, some interference power (IP) is likely always present at any given time, but it may occur quite randomly in location and time. While the capacity lost at any one site may not be large, aggregated across a network and viewed as a whole, the amount of lost capacity becomes quite large.

Being random and distributed, IP does not persistently exist at any one site for an extended period of time so a spot solution will not recover the lost capacity. Also, being distributed the value of spectrum conditioning is realized when aggregating the recovered lost capacity across the network versus for a single site. Our view of IP being distributed is based on the ISCO installed base and numerous analyses that have been conducted.

No one argues interference power is bad for a wireless network and the RF physical layer. RTWP or RSSI increase as IP is introduced, resulting in reduced coverage, diminished in-building penetration and less offered traffic being carried. ISCO assumes in the IVM that traffic is evenly distributed throughout a coverage area, so any reduction in cell radius or building penetration translates into lost capacity. Since fewer subscribers will be offered access, less traffic will be carried. The model refers to the amount of lost capacity in terms of reduced minutes of use (MOU), Erlangs and Mbps and revenue in US dollars. Once a dollar value is applied to these units of measure, a revenue figure can be calculated to reveal the value of spectrum conditioning. Moreover, with the wider deployment of spectrum conditioning, the cost of it decreases and the net economic gain increases.

To more accurately capture the true characteristics of a market's wireless network, the model assumes 12 different operating states based on 3 types of traffic loadings and 4 types of interference power scenarios. The base line used to determine the possible traffic that can be carried assumes zero interference power for each type of traffic loading. The model is then run 12 times to calculate the amount of traffic lost for each interference level and traffic loading. The aggregate of all 12 iterations of the model returns a total possible value that can be realized for a market's wireless network. The net value of spectrum conditioning is then calculated by subtracting the cost of deploying ISCO's Proteus from the revenue that was generated. As more Proteus units are deployed not only is more revenue generated, but the cost of spectrum conditioning goes down. As a result, net value increases.

Not considered in the IVM are the intangible benefits from the protection or insurance provided by the use of spectrum conditioning. From history we know IP will create physical layer impairments (PLIs) that will degrade service quality, KPIs, and data rates. The sources of unwanted RF are infinite and the types of interference are numerous ranging from cross border interference and co-site harmonics, to intermodulation distortion and the list goes on. However, placing a dollar amount on the value realized from these benefits is subjective. It has been proven time and time again that subscriber experience improves but quantifying the dollar value of that benefit is beyond the scope of the IVM.

There are three versions of the IVM, one for CDMA-EVDO and the other for UMTS. The IVM is built in Excel and by modifying the assumptions it can be tailored to any particular market.

The Impact

The differences between the voice and data traffic carried with zero IP and each particular combination represents the capacity that is recoverable. Then applying a dollar value to each unit of traffic will return a revenue value. The net value is the difference between the cost of spectrum conditioning and the recovered revenue.

For example, assume an average cell site during busy hour is engineered to support 87 Erlangs carrying 41,600 MOUs along with 16,560 Mbytes of data with no IP present. Based on model assumptions, the total annual revenue generated by that site would be $806,976. If we apply various levels of IP throughout the busy hour, the annual traffic carried will be reduced by 8% and corresponding revenue reduced to $742,418 for a single site. Since, however, we are not looking at one site in a vacuum the value will be aggregated across a larger portion of the network to become statistically valid. The aggregate total net value for spectrum conditioning illustrated for various levels of deployment is shown in numerous figures throughout the white paper.

Sign up as an ISCO Insider to receive a copy of the paper.

Maximizing Capacity Utilization by Minimizing Physical Layer Impairments
A Traffic-Based Model to Quantify the Value of Spectrum Conditioning

Proof of Value