DSS or SON. Which one is better in today's world?

In the past two years we have seen several companies claiming they have solutions for Self Organizing Networks (SON) which is supposedly a fully automated network optimization system. With this claims around I have not heard of any network that truly implemented a full SON. The most successful trials I heard so far are on automated neighbor list optimization, although this particular solution has been around for ages. So what's new? Has SON really arrived?

In the other hand there are existing Decision Support Solutions out there that can do neighbor list optimization, parameter optimization, configuration optimization and configuration management. Some of these solutions work and most works well on small networks. What about in big networks? Which is better today SON or DSS?

SON promoters claim 40% savings on OPEX, is this really true? Is it achievable? In SON, the idea is to minimize the lifecycle cost of running a network by eliminating manual configuration of equipment at the time of deployment, right through to dynamically optimizing radio network performance during operation. The idea is good but if the implementation so far only achieves neighbor optimization then the savings is virtually nil, zero, none, nada.... Why is that? Well neighbor optimization is available in most tools today and can be easily performed even in a non-automated environment by a skilled and experience engineer.

How about DSS? Is it any better? The answer lies on what the DSS solution is capable of doing. Is it capable of optimizing parameters, configurations, neighbor relationships, etc. If it is then there are savings to be had in terms of reducing the amount of human reliance and error while performing optimization tasks. However where is the greatest savings achieved?

My take is the greatest saving and impact is achieved by employing an end-to-end solution for monitoring, management and optimization. This can be done by using and correlating all performance sources in the network from radio to transmission to core. But how can this be done? There are several solutions in the market today and each has very different approaches. Some are solutions using single platform such as ActixONE (even claims full SON) or the Wireless Explorer. Some are piece meal solutions like the ones from NetCracker or PiWorks (no idea if it is). Each of this has their own ideas on how to reduce cost however it begs the question of whether it is a real cost savings or simply the same cost dressed differently.

When I say cost reductions I meant on all fronts from human resources to computing resources to management resources. How can this be achieved? I have fairly sound idea on how it can and if you would like me to share please drop me a line at ian.q@telia.com or follow my blog http://ianvernon.blogspot.com

Now do you know any solution out there that truly reduces the cost of operations? If there are let’s explore it and see if it stands the test of time.