Smart Grid: Multimodal Networks

What?

The industry has had some valuable experience with the first generation of smart metering and smart grids over the past few years. It is now becoming a familiar topic for most. The lifecycle of any technological solution evolves through distinct stages from its infancy, early years, middle years, late years, and finally to obsolescence. These new communication networks are just now evolving from the infancy years to the early years, so the internetworking solutions are still fairly young and in need of development and advancement.

The industry is now seeing these network designs move through a technological gate to the next stage - the second stage. I refer to this next generation of internetworking as multimodal networks. So, what is a multimodal network anyway?

I borrow the term multimodal from the transportation industry. In transportation, they have developed a slick standardized means to ship goods using various transports - trucks, trains, air, and sea - by using a common container format that nests perfectly into each of these methods of transportation. Almost any type of goods that can fit into the form factor of the shipping container can be moved around the globe in relative ease.

So What?

With smart grid networks, we have used stand alone, disparate networks to move datagrams from place to place. These discrete network tiers were oblivious to each other. They are able to crudely interconnect in order to move the traffic from one network tier to the next. But, these points of demarcation were clumsy and suffered from incongruence to each other.

Smart grid networks are like most networks in other industries and are designed with at least three network tiers:

• core network tier (optical fibre)
• distribution network tier [we call them backhaul network to avoid confusion with the distribution grid used in Utility parlance] (optical spurs, microwave spurs, WiMAX point to multipoint)
• access network tier (mesh or star point to multipoint)

In a modern smart grid, there can be more network tiers, segmentation of the main tiers, or other names used to describe the same tier. I have seen smart grids with 4, 5, 6, or even 7 tiers. For the sake of this discussion, let's just use the main three accepted tiers for internetworking.

Each network tier had its own technological approach so interfacing between these tiers often posed puzzling challenges for the network designer. Issues from these connection points included problems with:

• demodulation / modulation bridging in baseband
• different framing (TCP/IP versus Ethernet versus propriety frames)
• different ISO layers (Layer 2 versus Layer 3)
• VLAN compatibility issues
• QOS compatibility issues
• clocking / sync issues
• Security issues

With the next generation, the tipping point is now upon us beginning in 2015 and progressing into 2020, we are about to transition to a new approach to network design that models after the transportation approach - multimodal.

Multimodal networks join together in a seamless, transparent manner. With multimodal, we describe the different media used within the end to end network fabric. The magic of multimodal networks is the manner in which they interconnect. Let's call this the intermodal connection. Gone are the clumsy discrete inter-network interfaces with the listed incompatibilities. Now, we expect a tighter coupling between these network tiers to permit better, cleaner movement of the datagrams from one medium to the next. With this new transparent coupling, we anticipate:

• better performance with lower latency and higher data rates
• lower cost
• smaller more graceful designs for the network interfaces that we hang off our poles
• reduced complexity
• more features
• scalability and remote updates over the network fabric
• better control with a unified Network Management System (NMS)
• an easier transition away from a centralized architecture design towards a federated design that can "make and break" upon demand
• a standards based solution end to end

For example, let's consider the intermodal point where a WiMAX middle mile tier connects to an Internet of Things (IoT) last mile tier. In the second generation design, we look towards a more advanced standards based solution. So, the WiMAX would comply to the new WiGRID standards and the IoT would comply with the new WiSUN standards. The node where they meet would be a software defined radio perhaps making use of a suitable ARM processor with sufficient computing power to host both radios and excessive memory to add extra add-ins capabilities like a firewall. All of these functions would be in software rather than hardware. All of them would coexist within one power computational node that provided the two radio feature sets and permitted seamless flow of data both upstream and downstream between the middle and last mile tiers. The firewall with its IDS (intrusion detection systems) and IPS (intrusion prevention system) features would protect the network from hackers and crackers from entering. This is a transparent intermodal node.

Once all of these new capabilities are integrated into the multimodal network, we expect a new level of network availability and robustness that permits Utilities to operate a more reliable network for command, control, telemetry, and monitoring of the smart grid. The communication network needs to be far more robust than the power grid itself so it can survive outages and be used for restoration once the electrical system problems are resolved. This point is critical to the success of smart grids and is unfortunately often overlooked. That is, it is overlooked until the first major outage when operators realize that they are heavily dependant on the communication network for power grid restoration. If the communication networks are down too, then how do we restore the grid? Manually? Yes, we can do that, but then what's the point of smart grid? We learned this lesson the hard way here in Toronto during the ice storm of Christmas 2014, so heed our learning in your smart grid. The communication network must be available to restore the power grid! Therefore, anything that we can do to fortify the communication network is mission critical to the delivery of power, gas, or water to our customers.

What Next?

Beyond 2020, we can even predict greater advancements into 2025. Within the third stage changes to the network fabric, we look forward to an even more mature network that sees the multimodal / intermodal concept developed further. We hope to see developments such as:

• a software defined network (SDN) that separate the control plane away from the data plane. This allows centralized control with distributed data flows
• even more standards to drive deeper integration and smarter coupling between tiers
• an open architecture model that permits change and add-ins without problems or custom interfaces
• erasing the dependence on a single vendor solution, we want seamless interchange between vendors - think WiFi
• Harmonized end to end NMS
• an end to end federated network fabric

Of note is the level of federation that will advance with the third generation. Federation is the ability of the network to simultaneously exist as both a centralized and a distributed model with several tiers acting in harmony. These tiers can work completely independently or as one cohesive network fabric. Different tasks and functions of the network will normally reside on each tier. But, when circumstances demand it, these discrete functions can unify or decouple as necessary to solve the problems at hand. We expect to leverage federated networks with federated analytics, security, AAA, and much more.

Nowhere is this federation aspect needed more than in the network security. Today, likely 90% of all the Utility's data resides in the data centre. As we push smart meters and smart grid further and further, more than 50% of the Utility's data will reside on these network tiers. A lot of this data may never get back to the data centre as the intelligence is pushed outward towards the network edge with controllers housed at substation and closer to the end devices. New data or derived data may be all that is centralized in the third generation network. Regardless of where this data is located, it must be protected on the network edge just as it is in the data centre today. So, new thinking aimed at federating the security is required.

We are entering dark territory and voyaging to places that we have not been to before in network design. Sure, it is scary, but it is also a tremendous opportunity to bring new benefits in performance to the Utility and to establish higher standards for network characteristics aimed at enhanced reliability. In the end, Utilities will provide a more reliable service to their customers.

------------------MJM------------------

Michael Martin has more than 35 years of experience in broadband networks, optical fibre, wireless and digital communications technologies. He is a Senior Executive Consultant with IBM's Global Center of Excellence for Energy and Utilities. He was previously a founding partner and President of MICAN Communications and earlier was President of Comlink Systems Limited and Ensat Broadcast Services, Inc., both divisions of Cygnal Technologies Corporation. He holds three Masters level degrees, in business (MBA), communication (MA), and education (MEd). As well, he has diplomas and certifications in business, computer programming, internetworking, project management, media, photography, and communication technology.