By Jean-Paul Smets, Founder of Rapid.Space, a Hyper Open cloud provider based on open-source, open hardware, and
open service
In 1995, France-based
Alcatel-Alsthom was the number one telecom company in the world surpassing global players like Lucent, Nortel and
Ericsson. Today it no longer exists. Its facilities have been dismantled and
its remaining operations have been taken over by Nokia.
What
happened during 25 years to turn a
company at the forefront of telecommunications to one being forced to merge with a competitor? The answer
is simple: It refused to adopt new technologies. Alcatel was the industry
leader in telephone switchboards (PABX) but neglected to adopt technologies
which have today become industry standard.
What
happened to Alcatel and the PABX industry may also happen to the telephone
equipment industry. A technology invented eight years ago by Amarisoft, called Virtual
Radio Access Network (vRAN)
can provide better performance for 4G/5G networks at much lower cost than traditional hardware
from Nokia, Ericsson, or Huawei.
These
vendors provide classical base transceiver stations (BTS) which require their
dedicated hardware for handling radio signal baseband computing and transport
processing while vRAN allows using commodity PCs to do this. Connecting such a
standard PC to a Remote Radio Head (RRH) - the point mobile phones will connect to - means
4G/5G network access can be provided in a much simpler way and independent of
supplier hardware. It could turn into another iteration of "software is eating
the world" which Alcatel and PABX fell victim to - this time with two
initiatives called Open RAN and Simple RAN setting out to transform the telecom
infrastructure market as we know it.
What is Open RAN?
Open
RAN is an initiative that wants to open the interfaces and protocols within the
BTS. It introduces an accessible network element between the radio unit (RRH -
where the high frequency radio signal is digitized) and the distributed and
centralized units (DU/CU - where the baseband signal is produced). This allows
building systems from different hardware vendors. The standardisation of these
protocols is driven by a consortium of operators and vendors called the Open
RAN Policy Coalition.Their mission is to make sure that RAN interfaces are
interoperable.
However,
it remains to be seen whether this will be possible, because Open RAN usually
relies on Intel's FlexRan architecture, which partially offloads signal
processing into the RRH and partly accelerates baseband processing with Intel's
proprietary hardware. This means it is neither possible to use generic RRHs
which do only frequency modulation nor commodity PCs handling all DU/CU
processing. It is therefore difficult to imagine Open RAN being integrated into
existing commercial infrastructures, or even reusing existing RRHs, in its
current state.
In addition, Open RAN is deployed using a
combination of Ethernet, Common Public Radio Interfaces (eCPRI), and
Multiprotocol Label Switching (MPLS) which adds expensive switches and complex
network protocols, ultimately increasing the cost per antenna. The problem of
these additional costs is where Simple RAN hopes to be the solution.
Why Is
Simple RAN the better solution?
Driven primarily by Amarisoft, the inventor of
vRAN, Simple RAN tries to define a standard for vRAN based on a subset of the
architecture approved by the
3rd Generation Partnership Project (3GPP).
Simple RAN does not require any dedicated
hardware acceleration. All radio baseband signal processing can be handled by
Amarisoft's software on standard Intel, AMD, or ARM processors. Unlike Open
RAN, Simple RAN supports the "Split 8" approach, which means it is
compatible with almost all RRH suppliers. The other big advantage of Simple RAN
is that it can provide ‘all in one' 4G/5G base stations - even under an
open-source hardware license. This means that it can take only a matter of
minutes to deploy a 4G/5G network with a decentralized core network.
Simple RAN addresses the process of
standardizing these new protocols more openly. Its specifications
are public, which is not the case for Open RAN. The only requirement to adhere
to the Simple RAN initiative is to publish the information on how to interface any
two components - something that many telecom vendors still refrain from doing
in order to prevent competitors from gaining technical insights.
Comparing the pros and cons of both RAN
initiatives, one has the impression that Simple RAN offers the better
proposition for the future of the telecommunication industry albeit not being
supported by any of the major hardware vendors at this point. Open RAN is using
technology that is incompatible with current networks and is dependent on both
proprietary software and hardware from Intel, whereas Simple RAN is compatible
with existing networks, hardware independent, and only requires the software
from Amarisoft. If Amarisoft was to become open source, this would completely
tip the scale in the favor of Simple RAN and likely accelerate the telecom
hardware market's evolution from being controlled by monoliths to becoming a
more competitive, innovative and democractic ecosystem - with the last question
remaining being whether dedicated hardware for radio baseband signal processing
and its vendors will follow the footsteps of PABX.
##
ABOUT THE AUTHOR
Jean-Paul Smets, Founder of RapidSpace and initial author of ERP. Jean-Paul is currently in charge of RapidSpace international development, investment relations and technology partnerships. He graduated from Ecole Normale Supérieure with a PhD in computer science and from Ecole des Mines de Paris with a Master in Public Administration. He gained industrial experience in Apparel Industry, Oil industry, Non Profit Organizations and Lorraine Region Public Administration. He is an active member of Free Software associations and has played a key role in the Eurolinux campaign which succeeded in protecting innovation from software patents.