image of mouse being lowered above mousetrap


In the mobile IoT business it is rare to meet a customer that does not prioritize “range” as a purchase criteria. Our announcement of Haystack XR Mode is genereating enthusiastic response not only from users looking for better outdoor range over LoRa, but in indoor environments, too.

Out With Yesterday’s Indoor IoT Infrastructure

We designed DASH7 to provide an alternative to legacy “low power” wireless IoT protocols. Most of these were “indoor” home automation protocols that didn’t see much outdoor use and almost never saw mobile use. Because most were short-range 2.4 GHz protocols, they marketed a way for an endpoint to “mesh” a message back to a gateway using a type of mesh networking, which in practice was more marketing gimmick than functional.

But fast forward to 2019 and we are re-thinking the entire indoor connectivity puzzle:

Addressing low power mobile IoT use cases just requires a very different connectivity than those low power fixed use cases of yesterday. And the unpredictability of endpoint location is just the starting point.

LPWAN’s Matter For When Connectivity Really Counts

A convenient feature of fixed low power IoT implementations is that the endpoints are … fixed. This means during setup and testing of endpoints, the connectivity to a gateway is generally known and predictable. The smoke alarm on the ceiling either connects to the gateway or it doesn’t. If it doesn’t, the solution is either a) move the fixed gateway closer to the smoke alarm, b) add another fixed gateway nearer to the smoke alarm, or c) move the fixed smoke alarm/endpoint closer to a gateway.

Packet error may have many fathers …

In almost any mobile environment, however, the predictability of a successful message transmission is much lower than in a fixed environment. Outdoors, an endpoint can travel into a canyon or park behind a large metal truck — places even high power cellular often may not reliably reach. Indoors, an endpoint can disappear into a crowded parking garage or hide behind pallets of iron tools— even temporarily — and frustrate the ability to connect. In either case, assumptions about connectivity in mobile environments are extremely multivariate. And where there is variance, there is the opportunity for failure. And the opportunity for failure in our experience is positively correlated with unhappy users.

LPWAN’s like LoRa have much to offer in terms of better indoor signal propagation compared to the indoor radios of the past. All things being equal, LPWAN’s like LoRa will have fewer packet errors when connecting indoors but LPWAN’s — arranged in a star topology — may not reach every indoor location even when transmitting, even as we do at 60–100 milliwatts.

The answer to maximizing mobile indoor location lies in software.

Someone should build a LPWAN networking stack for this

But as we’ve demonstrated with XR Mode, connectivity is about more than a good LPWAN radio. A good LPWAN networking stack completes a mobile indoor connectivity strategy using:

  • Advanced Forward Error Correction. Reducing packet error rates is just as important to indoor connectivity as outdoor. We implemented LDPC forward error correction (FEC) — bypassing LoRa’s built-in FEC — with excellent results, improving signal propagation over LoRa by 10–20x. Today, Haystack is the sole implementer of such powerful error correction over LoRa.
  • Automatic Repeat Requests. DASH7 is a bi-directional networking stack that uniquely allows for intelligent message re-transmission over LoRa. If a gateway (or endpoint) receives a corrupted packet, it can tell the sender to try again. So simple, yet so hard to do for other networking stacks which may blindly re-transmit with no guarantee of message delivery.
  • P2P & Multi-hop. An endpoint hidden from view of a gateway is still discoverable by other mobile endpoints utilizing DASH7’s peer-to-peer and multi-hop capabilities. For example:
  1. Endpoint A is temporarily hidden in a storage room and obstructed from communicating with a gateway.
  2. Endpoint B is then triggered to search for missing Endpoint A utilizing DASH7 P2P comms.
  3. Upon detecting Endpoint A, Endpoint B may i) store the discovery event or ii) immediately retransmit (via DASH7 multi-hop) to the gateway.

A few things worth highlighting:

  • In this model, personnel, mobile assets, or mobile inventory may all serve to extend indoor connectivity to mobile endpoints that cannot otherwise be reached by a gateway.
  • There are two modes of connectivity using this approach: synchronous and asynchronous. Synchronous connections treat the second endpoint (Endpoint B above) as something akin to a mobile gateway and provides real-time visibility into Endpoint A. Asynchronous connections assume that both Endpoint A and Endpoint B are out of range of a gateway but Endpoint B will utilize store-and-forward to provide non-real-time visibility into Endpoint A.
  • Mobile Endpoint B can, if necessary, independently execute RSSI-based trilateration and calculate location of Endpoint A.
  • The same approach can be utilized for outdoor location.

Mobile Indoor IoT Is A Largely Untapped Market

LPWAN technologies like NB-IoT or LoRaWAN are optimized for fixed use cases like meter reading and are in no way optimized for mobility, indoor or outdoor. Haystack XR Mode simultaneously addresses indoor connectivity challenges faced by other technologies but also addresses LPWAN mobile indoor location challenges. While the “indoor-outdoor” IoT (things that move both indoors and outdoors) itself a significant opportunity, the mobile indoor market remains at the “pre-Copernicus” stage for everyone from large militaries down to small construction companies: there is still a great deal of guessing and assuming regarding the location and condition of the mobile things that matter to them.


Get the latest about Haystack by signing up for our newsletter here.