One touted advantage of 5G is the ability to use network slicing to create customized virtual networks featuring controlled parameters for network performance. That might logically have the highest value for new emerging applications that require ultra-low latency, such as unmanned aerial vehicle control networks, autonomous vehicle control or remote medical procedures. In the consumer space, cloud-based multi-player gaming might provide another example.
As a practical matter, one might point out that slow revenue growth in the mobile service provider business, plus attrition in legacy lines of business, puts a premium on discovery of new revenue drivers such as sale of network slices, edge computing revenue and internet of things revenue. On the other hand, operator revenue needs might not match up so clearly with customer desire to pay.
Though the potential network slice value proposition seems clear--customized networks optimized for lead use cases--the value-versus-cost case might be much less clear.
There are several reasons. As sometimes is the case, the 5G network might operate well enough in “best effort” mode that quality of service guarantees are unnecessary. With the growth of work from home and video conferencing, it is reasonable to suppose that we now have a big driver for better return bandwidth, something network architects have expected, are working on and which 5G supports.
The issue might be that as basic network performance improves, the demand for QoS features arguably diminishes, to an extent. For consumer or business video conferences, 50 Mbps upstream bandwidth might typically suffice, where upstream on 4G or fixed networks might not support even that level of typical return bandwidth.
To be sure, many potential enterprise use cases requiring low end-to-end latency or jitter performance should benefit from QoS levels of service. The point is that many use cases might well be supported by the best effort 5G network, augmented by edge computing, without necessarily requiring a network slice and the accompanying QoS guarantees.
In other cases, quality of service for some important performance parameter might be helpful, but not generally necessary. So in that case a premium-priced network slice perhaps offers too little incremental value for the incremental cost.
Proponents argue there is value in customized networks supporting employees and customers for business-related, industrial, medical and remote control applications that require communications with performance guarantees. The same might be noted for many online gaming, virtual reality or augmented reality use cases.
That is why many believe enterprise use cases will be among the key potential growth areas for 5G, though noting that faster broadband and fixed wireless are important in the consumer space. The point is that those new use cases must be discovered and popularized. They are not givens.
As always, there might be potential new opportunities for billing and charging. Decades ago, long-distance rates were set at lower levels for evening off-peak usage, for example. More recently, mobile operators eliminated the distinction between domestic local and long distance calling. That innovation arguably hastened the demise of fixed network voice.
The point is that, although in principle network slicing can provide differentiated quality of experience, the better “normal” performance of 5G might obviate those use cases.
Consider video conferencing. Since 5G supports upstream speeds in the 50 Mbps range for typical end users, it may not matter much that network slices might support guaranteed higher upstream speeds as a feature.
Network infrastructure suppliers and connectivity executives always are hopeful that new capabilities offering higher performance can be monetized. It typically is quite hard to realize those hopes.
No comments:
Post a Comment