ECD: What new microcontroller technologies are available to meet the growing demand for cloud connected embedded devices?

VISHAKHADATTA: For many connected embedded devices such as motion and light sensors in a smart home, the cost of installing new wiring to power these devices can be prohibitive compared to the cost of the device itself. As a result, these connected devices must offer superior power efficiency so they can operate by battery or harvested energy sources. To meet the power efficiency requirements of connected devices, MCUs must support ultra-low power consumption through mixed-signal integration. An on-chip DC/DC buck converter, for example, can enable higher energy efficiency while allowing MCUs to operate all the way down to the lowest usable battery voltage. A 32-bit MCU with an integrated DC/DC buck converter can achieve 40 percent lower active mode power compared to a similar MCU without a buck converter.

MCUs used in connected devices ideally must support multiple power domains, enabling peripherals to operate autonomously at different frequencies with the CPU powered down. For example, a Direct Memory Access (DMA) technique can be used to collect sensor data and wake the CPU when there is a full buffer of data to process. This results in a greater sleep-to-wake ratio and higher power efficiency. A 32-bit MCU with a dedicated Data Transfer Manager (DTM) hardware block can enable complex tasks to execute autonomously without CPU intervention. In these instances, the MCU core remains in its lowest power state until all tasks have completed. The DTM is useful in wireless data transfers in which raw data is processed through multiple operations before being delivered to the radio for transmission.

Connected embedded devices must be able to use wireless protocols such as ZigBee that are lightweight and offer data rates that reflect their requirements. The ideal RF technology depends on the specific application. For example, Wi-Fi is appropriate when high data rates are required for bandwidth-intensive tasks such as streaming video. For lower-bandwidth applications, 2.4 GHz ZigBee and sub-GHz RF technologies provide a more power-efficient wireless link that is easily integrated into the connected embedded device. For simple applications such as garage door openers or systems requiring long-distance connectivity such as smart meters, sub-GHz transceivers are an optimal choice. For mesh networks with a large number of connected devices, ZigBee technology offers a proven, robust solution.

A mesh topology is ideal for many connected device applications. Consider a home lighting system where the number of nodes can exceed 30 lights and sensors. Connected device applications based on ZigBee technology provide self-configuring, self-healing mesh connectivity that can interconnect hundreds or even thousands of devices on a single network.

Emerging sensor technologies are also enabling new capabilities for connected devices. For example, CMOS-based humidity sensors can be added to smart thermostats or security systems to provide accurate relative humidity measurements. Many advanced mixed-signal MCUs include integrated temperature sensors that provide exceptional accuracy over an extended temperature range, making them ideal for connected devices that require economical yet high-precision temperature sensing.

ECD: As you look ahead for the next few years, what trends do you see in the embedded design area and how will they affect your product development plans?

VISHAKHADATTA: The Internet has made huge leaps in recent years. IPv4 is giving way to IPv6 so that every connected device will have its own IP address. Machine-to-Machine (M2M) communication is enabling connected devices to exchange and act on information without the direct end-user involvement. This growing web of connected devices – the Internet of Things (IoT) – includes smartphones, tablets, TVs, gaming consoles, home appliances, security systems, smart meters, and many other devices.

Ultra-low-power MCUs, sub-GHz transceivers, ZigBee SoCs, and sensor networks will form the backbone of the IoT. Until recent years, these components were not power efficient, robust, or small enough to meet the requirements of connected devices. However, semiconductor technology has advanced rapidly to enable the development of energy-efficient end points that can operate for many years on battery power or even indefinitely on harvested energy.

Despite these advancements, connected embedded devices are not as ubiquitous or intelligent as they could be. By analogy, the car you owned 20 years ago conveyed basic information through simple dashboard indicator lights. Nowadays, cars feature sophisticated infotainment systems that provide detailed information and intelligence. A similar evolution is occurring with connected devices, and the key to this added intelligence is software. While embedded components provide the foundation for connectivity, application layer software enables the underlying M2M interactions that ensure that connected devices operate reliably regardless of operating environments. Developers can implement advanced functionality through software that extends the range of autonomous control to enhance efficiency and convenience. For example, while it’s helpful to be able to turn on a light remotely, it’s even more useful when the lighting system alerts us that an LED bulb needs replacing. A smart home’s sensor network could even determine when no one is around and power down all electronic devices. Such a simple innovation, multiplied over hundreds of millions of households, could save considerable energy.

ECD: With ubiquitous connectivity dominating embedded designs, what security precautions are available to prevent unauthorized access?

VISHAKHADATTA: Security for cloud-connected devices is a system-level challenge that must be addressed holistically through hardware and software. Semiconductor suppliers can help meet this challenge by embedding AES encryption blocks in MCUs and wireless devices. The end application generally determines the type of security technology required for an embedded product. For wireless products based on ZigBee or Bluetooth, security is built into the wireless protocol.

For most simple embedded applications, AES provides a secure, cost-effective solution that’s easy to implement in firmware or hardware. For low-bandwidth systems, such as a USB dongle, a simple firmware implementation of AES that uses a few kilobytes of code and minimal RAM can be effective. For higher-bandwidth systems, a hardware implementation of AES makes more sense, and there are many MCU products available with AES encryption peripherals. The benefit of choosing a flexible, commonly available encryption protocol like AES is that it does not define the hardware platform and lets engineers choose the right MCU for their embedded application.

Beyond AES, it is important for connected device developers to collaborate with software vendors and system integrators to tackle embedded security at the application layer.

ECD: Software development is a huge portion of each new embedded development project. What development tools and libraries do you offer developers?

VISHAKHADATTA: To streamline the development of embedded applications based on our 32-bit and 8-bit MCUs, Silicon Labs offers development platforms featuring interchangeable MCU and radio components and other subsystems.

Our complimentary Eclipse-based Precision32 IDE for 32-bit designs includes a compiler, debugger, and an online dashboard for application-critical information. A centerpiece of the IDE is our GUI-based AppBuilder software, which enables developers to graphically select their peripheral mix, optimize designs for ultra-low power, customize pinouts, and generate source code – all without having to write a line of code.

Our GUI-based Wireless Development Suite (WDS) software for EZRadioPRO transceivers enables developers to design and deploy sub-GHz wireless applications with little or no specific RF design experience.

For mesh networking applications based on ZigBee technology, we offer the EmberZNet PRO protocol stack, a comprehensive ZigBee development environment featuring visualization and debugging tools, and application templates for ZigBee Smart Energy, Home Automation, and Light Link profiles.

ECD: Does your company offer any educational events or online classes to help embedded designers get started with microcontroller-centric projects?

VISHAKHADATTA: We offer online video demonstrations and tutorials through our YouTube channel, http://youtube.com/viralsilabs. We also offer two-day ZigBee training classes in our Boston office each month. For in-depth customer education and training, we and our distributor partners offer training classes at classroom sites around the world.

Diwakar Vishakhadatta is the Vice President and GM of Silicon Labs’ Embedded System products group. He joined Silicon Labs in 1999 as the Wireless Director of Engineering prior to the divestiture of the company’s cellular business to ST- Ericsson (then NXP) in 2007, at which point he became Vice President of Entry Cellular Products for ST-Ericsson. Diwakar rejoined Silicon Labs in 2009 as the Director of Isolation and Power Products, and then became the General Manager of Broadcast Audio products in 2011. Prior to Silicon Labs, he served as a Design Manager at Cirrus Logic. Diwakar holds an MSEE from Oregon State University and a BS from the Indian Institute of Technology.

Silicon Labs www.silabs.com

Follow: Twitter Facebook Google+ LinkedIn YouTube