Sometimes radio paths that model as “perfect” are, in fact, perfect environments for multi-path cancellation.  Without the careful design of a network or the detailed planning of individual paths, you won’t be able to fully leverage your technology investment to achieve your network design goals and maximize your return on investment in spectrum & hardware.

John Wilkinson, Network Design Manager at Mimomax, gives his tips on how to optimize spectrum use and achieve better system performance through careful network design and path planning.


Tools, Datasets & Competency

Advanced software for designing wireless communication networks is available to help with radio network planning. The available tools range from capable freeware through to sophisticated packages, with the latter requiring a reasonably high financial investment. There is also a wide range of openly available terrain and land cover (clutter) datasets available. With the right knowledge and GIS tools these can be converted to the proprietary formats required for the planning tools. Commercial datasets can also be procured, however the cost of these can be considerable. Whatever tool is utilized, the old saying “garbage in – garbage out” applies, so having users that are knowledgeable and passionate about this line of work is imperative. Interpreting and clearly presenting the results, in conjunction with sound practical knowledge and field experience, is essential for optimal network design.

Commonly-used availability models such as Vigants-Barnett and the ITU-R P.530 series were mostly developed based on the performance of reference microwave paths, so caution needs to be exercised when applying these to lower frequency VHF/UHF paths. A predicted Fade Margin and associated Availability may look good on a desktop study, however these results can quickly be eroded by real-world issues, such as site noise/interference/blocking, unaccounted near site obstructions, higher than expected path loss or deep multipath fades.  An experienced network designer will have the skills to consider predicted results and assess these in conjunction with other tools such as Google Earth, site/path surveys and local knowledge.


Channelization of Spectrum

In a capacity constrained environment, the ability to channelize and reuse your valuable spectrum assignment becomes a high priority.  Assigning narrowband channels that can be deployed with spectrally efficient radio technology (up to 16 b/s/Hz with MiMOMax) is key to developing a channel plan that supports a wide area PtMP deployment.

When sizing network capacity it is important to keep the future in mind. It is unsurprising that future predictions tend to support an increase in data requirements to the edge.  The use of technologies such as MIMO, full duplex and high order modulation will help to push more data through the channel.


Antenna Height Optimization

Height analysis can be carried out in various tools which will identify if a path will be susceptible to antenna height issues.  However, it makes sense to a) not rely solely on the data produced by the desk model and b) to check real-world results after antenna installation.  While it sounds counter-intuitive, starting with the antenna high at a site may not yield the best performance.  In some multi-path sensitive environments, dropping the antenna can actually achieve better results.


Paths over Water, Marshes or Deserts

With tidal paths experiencing height variations of 2-5m and deserts or marshes potentially becoming lakes in the rainy season, paying close attention to the subtleties and seasonal changes of each path is crucial.  Reflection off water or other flat surfaces, such as sandy deserts, may require an approach incorporating spatial or frequency diversity to increase your availability stats.  An alternative to this approach, however, is to model different antenna heights.  Achieving the right height may effectively accommodate tidal effects, without the need for diversity.

Some of the more powerful modelling tools on the market can give a highly detailed analysis of sea height tidal changes and the impact of weather conditions on K factor (curvature of the earth) and the resulting effect on propagation.   Depending on the terrain at your sites, investment in a more expensive modelling tool could pay dividends when it comes to achieving high rates of availability.


Real World Experience (knowledge)

Beyond the analysis of data generated by prediction tools, information gained from the real and virtually modelled world can be invaluable when it comes to the assessing accuracy of predictions.   What to consider:


Google Earth

The better network design tools available can export coverage prediction results and 2D/3D Point-to-Point path models in Google Earth format. Analysis of the results in Google Earth is an invaluable tool for visualizing results and identifying potential issues. The resolution and accuracy of terrain, imagery and 3D building layers in Google Earth is outstanding and adds immense value to the network design process.  Objects that are not captured in 3D in Google Earth, or dynamic objects such as trucks on a roadway that have the potential to obstruct a radio path, can be simply modeled via manual addition of extruded polygons within Google Earth. Static objects identified in Google Earth can also be added to the clutter database of the network design tool being utilized.


Geographic variations

Geographic considerations will also need to be incorporated – assumptions on tree heights, for example, will vary greatly in different regions of the world based on average rainfall and growth rates.  Manually adjusting clutter and terrain data to suit the site locality will offer far greater accuracy in predictions of link availability.


Site Surveys & Drive testing

While you can make preliminary predictions using desk-based models, it is during site surveys and drive testing that new discoveries may be made. These discoveries can be used to enhance clutter data (i.e. tree heights) resulting in changes required in antenna heights or directions.  Feeding this new data back into the datasets will also positively impact the accuracy of future network expansions.



In conclusion, a raft of different tools, prediction models and datasets exist to assist in radio network design.  However, key requirements for success are:

  1. An experienced and dedicated operator to capture and enter the network requirements;
  2. Accurate setup of the tool to generate meaningful results for analysis;
  3. Smart utilization of spectrum, channelization and clever frequency use plans to minimize interference and ducting, and;
  4. Drive testing and site surveys to ensure real-world experience complements the data created by desk-based modelling.