The Internet of Things (IoT) shows no signs of short-lived popularity, with more and more interconnected things expected to appear regularly over a decade. This is not surprising, as we are constantly imagining finding new ways to use technology to do things, and developing entirely new applications of old and new technologies. And each new application adds many endpoints to the network, making the IoT increasingly larger.

By Julio Song, ON semiconductor Logic, Switch and Storage Business Unit

The Internet of Things (IoT) shows no signs of short-lived popularity, with more and more interconnected things expected to appear regularly over a decade. This is not surprising, as we are constantly imagining finding new ways to use technology to do things, and developing entirely new applications of old and new technologies. And each new application adds many endpoints to the network, making the IoT increasingly larger.

This trend is evident from the proliferation of different wireless technologies currently in use that operate in most of the RF spectrum. They are mostly included in unlicensed frequency bands under the Industrial, Scientific and Medical (ISM) definition, usually excluding applications that might be classified as telecommunications.

make life more comfortable

The large number of IoT endpoints involved can be in the tens of billions, if not hundreds of billions, making debugging and maintenance challenging. Assuming we don’t create a huge waste management problem for the next generation, these devices will all be running for years and thus need to be designed with maintenance and repair in mind.

Complicating this logistical issue is that many IoT endpoints are very low cost and the level of maintenance they provide may not support the relatively high cost of on-site maintenance. Infrastructure costs must also be kept as low as possible, so license-free wireless solutions are used.

Even relatively small factories may soon have thousands of IoT endpoints, so it’s not hard to see why mesh networking technology has become the preferred local area network topology. It enables remote monitoring and maintenance of each device, and also provides a degree of redundancy in the network, as the mesh topology can withstand those major outages that might be caused by radio interference or the total failure of some endpoints.

Another popular topology in IoT, especially for sensors located in remote areas, is the star network employed by Low Power Wide Area Networks (LPWANs). These networks prioritize range and low power over payload, often supporting very short transmissions over long distances, perhaps limited to once a day. Therefore, they are generally best for work data that is not affected by latency.

The choice of network topology will almost entirely depend on the application, and although the financial cost of the endpoint may be relatively low, it may provide significant data. Some analysts estimate that LPWAN will be worth $50 billion in just a few years, a figure based on the total value of devices and the services they provide.

When an endpoint in the field fails, it almost inevitably needs to be repaired. If it is not possible to do it remotely over the network, it requires out-of-vehicle repairs, which requires a service engineer to go to the site and solve the problem as quickly and efficiently as possible.

Once on-site, the problem may increase because there is a high chance that the endpoint will not be physically or electronically accessible. Even if the endpoint is easily accessible, engineers can have a hard time figuring out why it isn’t working because the endpoint’s design may be based on a single system-on-a-chip (SoC). The current level of integration makes this more and more common.

Two ‘E’s for IoT

All areas served by the electronics industry store data about device activity locally, which has been around longer than the Internet of Things. Logging data using non-volatile memory has been established as an efficient solution for recording critical system information. Data stored in non-volatile memory can give you insight into how your device is functioning and why it stopped working. They are also widely used to store functional parameters, calibration data, and other types of information that may need to be updated periodically but can be retained in the event of a power outage.

This activity puts enormous pressure on the underlying storage technology, as it requires extreme endurance, higher than that offered by some popular memories such as flash memory. Therefore, the preferred technology is EEPROM.

In some applications, it is not uncommon to write to EEPROM continuously during normal operation, driving the need for memory technologies that can withstand millions of read and write cycles without failure. This does not include other types of memory that may only be able to support program and erase durability of 100,000 cycles or less.

Now, by integrating EEPROM storage and RFID connectivity in a single device, engineers can design IoT endpoints to store operational data and communicate with service engineers even in the event of a power failure or complete power outage. This presents an entirely new paradigm for service and maintenance, and is fully applicable to IoT (see Figure 1).

Why RFID plus EEPROM equals simpler IoT

Tablet/Phone: Tablet PC/Phone

RF Reader/Writer: RF Reader

LED TV: LED TV

Motor Trim from EEPROM: EEPROM adjusts the motor

MCU: Micro Control Unit

Adjustable Current Regulator: Adjustable current regulator

TV Trim from EEPROM: EEPROM trim TV

Trim parameters, diagnostic data, remote maintenance: tuning parameters, diagnostic data, remote maintenance

Figure 1: Adding RFID-enabled EEPROMs to IoT endpoints will provide a new dimension to debugging, repairing, and maintaining billions of devices

Perhaps more importantly, the solution in question extends the operational distance of the wireless link from less than 10 cm (a typical value for passive RFID) to 150 cm. This exactly takes the application of passive RFID in IoT to a new dimension. The ability to use RFID to read and write data to EEPROM over a distance of 1.5m, even when the system is not powered on, will help engineers more cost-effectively debug, maintain, repair and repair endpoints after they are deployed in the field.

unique solution

The N24RF series of RF EEPROMs developed by ON semiconductor integrate an ISO 15693/ISO 18000-3 Mode 1 compliant RF transceiver with 4, 16 or 64 kbit EEPROM memory in 8-pin SOIC or TSSOP packages. It offers 2 million program, erase cycles, 200 years of data retention, and operates over a temperature range of -40 to +105°C.

The device uses passive RFID and therefore does not rely on an external power source. Instead, it gets all of its power when connected to an external coil antenna. The device is classified as high frequency (HF) RFID and operates at a carrier frequency of 13.56 MHz, enabling it to communicate with RFID readers at low speed (1.65 kbit/s) and high speed (26.48 kbit/s), up to Fast instructions up to 53 kbit/s. It is a unique method of passive RFID implementation that achieves this at a distance of 1.5m.

Using passive RFID, IoT endpoints can be interrogated to recover important data records after a failure, even if the rest of the circuit loses power or fails. It also supports over-the-air (OTA) updates of calibration or operational parameters while the endpoint is still running. Inter-chip communication is implemented using the I2C bus, and the host processor can read and write from the device during normal operation, allowing calibration or operating parameters to be updated in the field without service interruption.

The N24RFxx devices use Reader Talks First (RTF) technology, which wakes up when an electromagnetic field is applied via inductive coupling. It provides a wider range, meaning engineers can query potentially inaccessible IoT endpoints, such as light fixtures, by simply using an RFID reader located beneath the light fixture.

Security features include a 64-bit Unique Identifier (UID), as well as support for multiple 32-bit passwords, with locking for different storage sectors. Another feature (in selected devices) is a voltage output pin that provides enough power to support a standalone ultra-low-power microcontroller.

Summarize

EEPROMs are widely used for data logging and parameter storage in many applications that require high endurance and proven data retention. By adding RFID capabilities, the same data can be accessed wirelessly and securely at a distance of 1.5 meters, providing a new dimension in IoT endpoint design.

The Links:   LC171W03-A4 AT070TN07-VD