markrubens - Fotolia
It's that time of year again. It's time to look back and forward at the networking industry and, in particular, my area of expertise: wireless networking technology. This year, like past years, I have a few predictions. But I'd also like to focus on a wish list for trends in wireless networking technology.
As a technology analyst, I pore over voluminous amounts of data and synthesize a rational and justifiable prediction of the future. I also point out what's missing amid wireless networking trends.
As an example from my early days in covering wireless networking technology, I wanted an inexpensive, hand-held, unlicensed-band spectrum analyzer. A couple of years later, I received one. Today, spectrum analyzers remain the best way to diagnose issues in Wi-Fi and other wireless systems at the physical level. These devices are essential for any organization's or service provider's wireless toolkit.
The same could be said for centralized control and management of access points (APs), wireless analytics, AI, machine learning and many other technologies. Ask -- or maybe predict with a high level of certitude -- and you shall receive.
So, here we go -- requests and predictions for trends in wireless networking in 2019:
1. Millimeter waves
Frequencies above 30 GHz -- and especially the big swatch of unlicensed spectrum around 60 GHz -- remain underused today, which is a shame because of the vast potential here. Component prices have fallen, and ease of use has improved. I still have high hopes for 802.11ad and its successor, .11ay, which could result in 100 Gbps wireless LANs.
But, expect to see a bunch of fixed point-to-point, point-to-multipoint and mesh services next year and eventually 5G mobile services. Spectrum is the highway for wireless, and we can't afford to ignore any roads, as traffic volumes continue to grow.
2. Built-in performance evaluation and management tools
This one drives me crazy. I want to measure the performance of a given link between a Wi-Fi client and an AP, but there's no way to do so. I need a benchmarking server built into the AP to eliminate the need to go through the AP, across another link and into a server, thereby introducing more variables. This seems like a no-brainer, and yet it remains to be invented. I hereby put all patent rights in the public domain and have my credit card at the ready.
3. System-independent, edge-to-core analytics
Analytics, increasingly based on AI and machine learning, are essential. Networks today are simply too complex for mere mortals to understand, configure, optimize and troubleshoot, while also maintaining ROI. We need analytics that run across the entire network and across products from multiple vendors. In other words, we need end-to-end analytics. But this is a tough nut to crack, just like cross-vendor unified management has been. However, we need it, and someone is going to do very well here over the next few years.
4. One-stop wireless security as a service
No one is ever done with security; it's always evolving. And considering the range of potential threats will remain vast, an end-to-end cloud service could be the best way to address this challenge. We're making progress here, but, again, we're never done. Therefore, my wish for 2019 is simpler: May every organization everywhere have a solid, published and supported security policy. I'm amazed at how many organizations do not.
5. Carrier-independent services
I know no one wants to be the big, dumb pipe, but someone has to do it. Ideally, several someones would do it. A competitive market boosts performance, reliability and availability, while driving down costs. Sure, the carriers will continue to push into value-add, and that's fine. But all services ultimately need to be independent of carriers.
Think about it: You buy broadband access from a carrier, which is a data plan with an IP address, but your phone number -- a service on top of IP -- is elsewhere. The same could be said for multiple mobile personas, unified communications, and just about every other networking and communications service.
6. 'Get smart' replaces 'get more' for network performance issues
Historically, the strategy for dealing with performance issues, such as system-wide capacity, has been to upgrade to the next generation of technology -- like 10, 100 or 1,000 Gigabit Ethernet and 802.11b/g/n/ac/ax. And that's fine, but the more important trend going forward will be to apply more intelligence and not more brute force. Consider, for example, AI, machine learning, software-defined anything and network functions virtualization.
Over time, networks are predominantly implemented in software, with less specialized hardware. These networks should have more flexibility and adaptability thanks to software running on high-performance, inexpensive processors. We're already seeing, for example, key elements of 5G networks implemented as software running on commercial off-the-shelf computers and open source operating systems. And we'll see a lot more across essentially all networks, everywhere.
Tracking emerging trends in wireless networking is an exciting corner of networking that's always innovative, complex, fast-changing and, most importantly, fun. I bid you all the very best for your wireless networking endeavors in the new year.