The Spirant two focuses an increased focus on the AI test in the telecom run by the two ‘mega-trends’.
Rapid growth in AI is affecting testing ideas in many ways. Recently at RCR Wireless News Webinar, Stephen Douglas, head of Spirant Communications of Market Strategy, underlined some factors in the fast changing landscape.
Douglas made changes in the context of two mega-trends.
The first is Network is being increased with Artificial Intelligence Tools“We are currently looking at, many sellers of network devices begin to start bringing in AI capabilities in their solutions – whether it is in switch and router, whether it is in radio equipment, firewall, gateway, at setter too Ho, even the core network, “Douglas explained that AI is being used, for example, to support things like dynamic policy configuration For and watching telcoses, how it can be embedded in load balance, increasing energy efficiency, mobility adaptation and so on in radio access networks.
Douglas said, “We are looking at a fleet of AI coming in the network in this way.”
How does this test affect? The efficacy of that AI is to be confirmed, both before deploying and once active in the network. Douglas has summarized some questions that have to respond to the testing of the AI network tool: First, is AI working? What are the results are better or better than old, non-AI-based systems? Is it providing benefits, or AI is introducing new risks?
The second is mega-trend Network being created to support the use of AI And its demand for calculation power, bandwidth, delay and so on. Data centers are already feeling these effects and supporting GPU groups in terms of power requires changes in their design and architecture, but Douglas said that traffic behavior and performance within the data centers also demand Changing with AI and they are also affecting changes. The amount of wireless and wireless network AI traffic increases. (Chat alone, it should be noted, hit a milestone of 400 million active weekly users this week.)
As the service provider sees to adapt and upgrade your network to support AI traffic, they will also have to test for the parameters that demonstrate whether their network less delay, defects, throoputs and specific performance characteristics Can provide AI workloads that can provide.
Douglas identified three capabilities, enabling AI testing, both in the context of the use of AI in the network, and the ability to support the network and support AI in the network. they are:
-Using exemplary replicas of the network as playgrounds to test digital twins, or AI.
-Cynthetic test data, realistic but simulation data using either to fuel AI training, or test the system.
-Condin and active testing -Non only in the laboratory environment, Douglas reported, but within the live network.
He offered some examples: When creating new data centers for A architecting and AI support, Douglas said, “By this time, the only way to test it to use real expiry, real GPU is that. In fact, the real GPU is to use. Stimulate the load. And it is very expensive, a lot of time taking and not very realistic. “But now, a digital twin can be used to generate different types of traffic and behavior without overhead and the cost of using real calculations expon.
Douglas also said that digital twins are being used for safety testing, realistic attacks to look at the effects of traffic and loss, as well as to check that AI-equipped firewalls are actually treating traffic The way they are considered.
For more insight on AI tests and trends, look at the demand for full webinar, which is characterized by Spirant Communications, Viavi Solutions and Drivats, and read more of RCR Wireless News’ AI coverage.
