Page 16 - Security Today, November/December 2020

“As the well-worn adage says, “Necessity is the mother of invention.” The COVID-19 pandemic is no exception. During 2020, many IP devices—originally designed for security--have been integrated with other technologies and put into service in new ways.”
cessed. Network micro-segmentation lets businesses apply vary- ing levels of security to specific parts of the network where more critical data resides. Granular network perimeter security – based on users and devices, their physical locations, and other identify- ing data – determines whether their credentials can be trusted to access the network.
Layered cybersecurity. Another trend gaining traction is hard- ening a device’s operating core where sensitive information such as end-user data, log-in credentials and network access certificates reside. In the coming year, I expect more manufacturers will be in- corporating chipsets encoded with protective features like device identification to prevent port hijacking, secure boot and signed firmware checking to prevent unauthorized or malicious down- loads. They’ll be implementing a more layered approach to cyber- security so that each layer of hardware and software shields the layer below with additional security features. To further reduce cy- ber vulnerability, we’ll see more manufacturers automating proce- dures to ensure that customers implement changes, upgrades, and patches throughout their ecosystems in a timely manner.
Device management and lifecycle management. Two other aspects of cybersecurity coming to the forefront are individual device management and comprehensive lifecycle management of the entire surveillance solution. With cyberattacks becoming in- creasingly numerous and sophisticated, it’s becoming even more important for manufacturers, integrators and end users to work together to institute consistent hardware, software and user poli- cies that conform to cybersecurity best practices and continue over the lifetime of the ecosystem and its individual components.
Fewer OEM relationships. OEM partnerships will become less common as it is difficult for the end-user to know the real manu- facturer, which is a must for valid cybersecurity processes.
SHIFTING MORE COMPUTING TO THE EDGE
Billions of devices are already connected to the network and this number continues to increase exponentially. With all that potential computing power at the edge, there’s more momentum for manufacturers to harness that in-camera resource to help cus- tomers detect and identify threats, and make split-second deci- sions on how to mitigate them.
Driving the move to edge-based processing is the opportunity for customers to reduce system complexity and lower their operating costs. They can build faster and more scalable systems, and facilitate more complex, real-time analysis of events at the point at which they occur, which can lead to more timely response to incidents.
In the coming year we’ll likely see more powerful edge devices embedded with deep-learning processing units (DLPU). These devices will provide a perfect platform for third-party developers to create analytics based on AI and deep learning that can provide more granular object classification and more predictive behavior
analysis. Furthermore, because these deep learning programs only transfer relevant video over the network, customers will be able to lower bandwidth consumption and reduce their storage needs. With the improvement in analytics accuracy, customer will also experience fewer false alarms which will lead to additional cost savings and more efficient use of human resources.
REACHING FOR MORE PREDICTIVE ANALYTICS
Inspired by the way the human brain processes information, manufacturers will continue the trend of building cameras em- bedded with deep learning processing units. Developers will be able to use these DLPUs to improve the reliability of their video analytics algorithms and enhance surveillance, security and busi- ness processes.
These deep learning tools correlate the massive amount of data collected by devices and sensors to train the analytics to bet- ter recognize patterns, classify information and make decisions by labeling and categorizing what the camera sees. While it will still take some discernment on the part of customers and integra- tors to separate hype from fact, the opportunities for deep learn- ing solutions will rise as the technology matures. With contin- ued investment in refining this learning process, I expect we’ll see dramatic increases in video analytics accuracy, and a demand for these solutions, in the coming years.
BRIDGING THE EDGE AND CLOUD COMPUTING DIVIDE
As we look towards 2021, with more processing power avail- able in edge devices, hybrid computing solutions will become an emerging trend. Companies will be looking to implement smart architecture that directs processing tasks across the edge, core, and cloud to wherever it makes the most sense. These multi-tiered solutions will give customers the flexibility to deploy and manage their ecosystem more efficiently and more cost-effectively.
The traditional client-server model alone presents limitations, inefficiencies and network performance issues due to centralized cloud servers that must respond to a multitude of device requests. Leveraging a hybrid edge-cloud model—where central and edge computing resources are combined—can reduce cloud hosting costs, bandwidth and latency issues. With more powerful devices companies will be able to take better advantage of processing at the core, the cloud and now, the edge—essentially where another server can process significant amounts of data on the device.
As an added benefit, edge devices offer an unparalleled lev- el of built-in cybersecurity features. Many of these devices can be managed with on-premise tools that allow users to perform device configuration, back up, firmware upgrades and manage cybersecurity controls—ensuring a greater level of security than cloud environments.
MITIGATING RISKS AND CAPITALIZING ON OPPORTUNITIES
As in past years, the security industry will have its hands full mitigating the risks of cyberattacks, navigating regulations on pri- vacy, information collection and storage, and meeting customer de- mands for tools that can help them be more proactive in protecting their people, property, and electronic data. But
with each challenge there comes new opportu- nity to devise new technologies and solutions to create a smarter, safer ecosystem that can better meet customers’ security and business needs.
Dr. Martin Gren is the co-founder of Axis Communications AB.
16
NOVEMBER/DECEMBER 2020 | SECURITY TODAY
2021 TRENDS