Federal agencies face a pressing need to modernize their network infrastructures to support the growing demands of AI-driven operations and data-centric citizen services. With the increasing number of connected devices and the need for higher data capacity and network speeds, legacy networks are falling short.
MeriTalk recently sat down with John McDonald, senior director, Federal, RUCKUS Networks, and Surya Durvasula, vice president and general manager, Dell Technologies, Federal, to discuss how agencies must adopt a comprehensive approach that addresses all layers of the network, from the physical layer to the application layer, to overcome this massive challenge.
MeriTalk: Recent research from RUCKUS and Dell Technologies found that 82 percent of Federal government decision-makers said their network impedes their agency’s ability to meet evolving mission demands. What are the most impactful mission demands that aren’t being met by current agency networks, and how are agencies working to improve networks to meet those demands?
McDonald: While trying to take advantage of emerging technologies like machine learning, artificial intelligence (AI), and automation, agencies face exploding amounts of data from an exploding number of devices at the edge of networks. From facial recognition at the border to mail processing, agencies are generating massive amounts of data. At the same time, many are operating on legacy networks that are not designed to move these massive data volumes.
Start with the basics. If the physical medium for transmitting data is outdated or insufficient, the network can’t possibly satisfy the speeds and throughput needed to move massive amounts of data. The baseline of the network needs to be ready.
Durvasula: Quite frankly, the buzz around AI and Gen AI, and how they are going help agencies improve citizen engagement, is really helping agencies see that they have a bigger problem on their hands. Many times, agency decision-makers are thinking about replacing technology with more of the same or something similar. Now, we’re seeing some agencies looking at how they can create faster connections, closer to the edge, to ingest data faster. It’s a heavy lift, and they have to think about things differently.
Where we have sensors at the edge, why not bring the AI to the edge and use a really low-latency network to bring relevant pieces of information back to be processed versus trying to move massive amounts of data back and forth?
If agencies can manage data at the collection point, they’re reducing latency, improving bandwidth, and reducing the amount of classification or tagging and data retention because they haven’t brought that data back into a government facility. We help agencies accomplish this with Dell Federated Learning capabilities, which process data at the edge, giving agencies the insight they need to make faster decisions.
MeriTalk: As edge computing and the Internet of Things extend networks, and AI and machine learning exponentially increase data transmission, what are agencies prioritizing to address and manage network complexity?
McDonald: We’re seeing all of these factors increase network complexity, plus the government is also trying to respond to employee expectations for fast, robust connections anywhere. When employees worked remotely during the pandemic, they had better IT systems at home. As they return to the office, the government is trying to meet expectations for speed and efficiency with advanced Wi-Fi, LTE, 4G, or 5G services.
As the government tackles this complexity, we’re seeing more as-a-service models like infrastructure-as-a-service and network-as-a-service. The Department of Defense is evaluating this approach. By outsourcing network operations, valuable government talent can focus on analyzing data to derive insights for better decision-making in areas like national security and citizen services.
MeriTalk: Demanding modern workloads require high-performance networks that can adapt to changing demands. How can open networking enable optimized network performance for Federal agencies?
Durvasula: I think open networking gives us a lot of opportunity to bring relevant technologies into the ecosystem that meet the demands and requirements of the specific use case. Think of sensor data, medical devices, and kiosks. These devices are at the edge, but they’re three completely different use cases. Having an open network with the flexibility to adapt to specific needs is a major benefit. And as data grows, agencies have the scalability they need – and also choice and cost efficiency. This is an exciting moment in the networking space.
McDonald: RUCKUS’ heritage has always been about non-proprietary, interoperable solutions. That gives you options. You can build best-of-breed networks and put best-of-breed software on top of these networks. Competition breeds better use of valuable government dollars and breeds innovation.
The military uses a planning mechanism called PACE, which is primary, alternate, contingency, and emergency. You can apply that idea to networks, where an agency could use different vendors that can interoperate within those four segments. The idea is to not be locked into one solution that can be a huge security or supply chain risk.
RUCKUS has provided fully Federal government-certified wired networking for more than 30 years. We’re only one of three Federally certified Wi-Fi vendors. The non-proprietary nature of the solutions that we build allows us to partner with companies like Dell to offer more robust solutions to agencies. Working together, we’re doing really innovative things with data storage and data analysis. We help Dell’s customers seamlessly move that data around where it needs to go from a networking perspective.
MeriTalk: How does RUCKUS enable Federal organizations to update their network infrastructure to accelerate modernization efforts?
McDonald: The journey is going to be different for each agency and each bespoke application. For example, if an agency plans to use a Wi-Fi 7 or 5G network, it needs to consider the networking components at every level of the Open Systems Interconnection (OSI) model. RUCKUS can help agencies update network infrastructure based on their specific needs for use case development and application modernization.
McDonald: It is. Let’s say we’re building a robust Wi-Fi 7 network. We have to start at Layer 1 and work our way up. We need a multi-gig switch and cable that can support higher data rates and higher Power over Ethernet standards to deliver more power to edge devices. We also need laptops and mobile phones that can operate on Wi-Fi 7. It’s very challenging for vendors, because they want to sell the technology that they produce. But to enable that technology, the underlying infrastructure may need to be modernized first.
MeriTalk: What challenges do you expect Federal decision-makers to navigate in the next couple of years? How can network modernization provide a foundation for responding to these challenges?
Durvasula: Modernized network infrastructure is essential for cybersecurity and AI. Cybersecurity requires advanced tools, data management, and a strong, resilient network foundation. AI’s success depends on addressing data management and infrastructure challenges. Ultimately, if we’re not looking at this as a real opportunity to modernize the infrastructure, we’re going to have shortcomings when we implement the technology. For example, we can upskill the workforce to take advantage of AI, but if the service underneath isn’t modernized, AI is potentially not going to provide its intended value.
McDonald: The government procurement system wasn’t designed for the fast pace of innovation, and budgets aren’t unlimited, so the biggest challenge is determining what to do first – what will affect the government mission most effectively. Many agencies will need to work their way backward from Layer 7 to Layer 1. focusing on what’s needed to support their data rates and workloads.