Inter-comparison of Microwave and IP Radio technologies

Below, there is an inter-comparison of Microwave and IP Radio technologies and how each affects end-user voice quality.

Microwave Radios

Microwave Radios, mostly using FDD (frequency division duplex), often have Ethernet and/or TDM service interfaces such as E1s or STM (most often STM-1).

• Ethernet interfaces on microwave radios can be 10BaseT, 10/100BaseTx (10/100BaseTx being the most common of the 10/100BaseT family and also referred to as Fast Ethernet or FE) or 10/100/1000BaseT (also referred to as Gigabit Ethernet or GbE).
• The TDM E1 interfaces are electrical 120? (balanced) or 75 ? (unbalanced).
• STM interfaces are electrical or optical (often provided using SFP modules).

Microwave Radios use FDD for wireless access, which means that there is guaranteed throughput capacity for both transmit and receive directions.   Furthermore, low (compared to IP Radios) and constant latency is guaranteed.  For TDM, the latency across a link is often less than 500 micro seconds.  The timing jitter across microwave radio links can be measured against the ITU G.823 (‘Jitter and Wander’) specification.

The management of microwave radios is often done ‘Out-of-Band’ (OOB) using web-based interfaces, a CLI or SNMP.   OOB offers additional security because the management channel is physically independent of the ‘service channel’.  The ‘service channel’ transferring ‘customer data’ conveys voice, video and other data traffic either via transparent (‘native’) Ethernet, E1 or STM.

IP Radios

‘IP Radios’, most often using ‘shared-access’ TDD (Time Division Duplex) for the radio-interface.  These radios work ‘outdoors’ in the 5GHz license-free bands and are fitted with 10/100BaseTx or 10/100/1000BaseT Ethernet Interfaces.

IP Radio links often have much higher and variable latency (e.g. 2msec ~ 20msec+) – most often they work on an 802.11-specified ‘shared access’ basis and must co-exist with similar radios (utilization efficiency varies as the number of ‘other’ radios changes).

To give an example considering standard 2×2 MIMO 802.11n equipment, such radios can be configured to automatically choose between the best-performing of 16 MCS (Modulation and Coding schemes), 8 of which are single spatial streams and the other 8 using 2 spatial streams.  Such an IP radio can choose between 4 modulation schemes (BPSK, QPSK, 16QAM and 64QAM).   BPSK is transferred using one coding rate (1/2), QPSK and 16QAM use 2 coding rates (1/2, ¾) while 64QAM can use 3 coding rates (2/3, ¾ or 5/6).

So, why is there higher and more variable latency for an IP Radio than for a Microwave Radio?

Contributors to variable latency and data throughput speed for IP radios include:

• the addition of 802.11 radio MAC-interface overhead,
  • Depending on if standard IEEE802.11 or WDS is used to bridge Ethernet data across a radio link, the number of MAC addresses transferred back and forth across the radio interface is 3 or 4 respectively.
• the delay due to channel access control mechanisms
  • checking if the channel is free and dealing with channel contention (CSMA/CA): variation in channel parameters such DIFS and CW_min and the number of channel users affect overall channel throughput capacity,
  • catering for the presence of legacy 802.11 client devices,
  • system trade-offs between diversity (reliability) and multiplexing (for higher throughput) as the radio links automatically attempt to optimise setting of link parameters in order to achieve maximum reliable throughput,
  • the need to wait for frame receipt Acknowledgments and
  • Some IP radio systems are affected by other (interfering) systems that automatically change their frequencies (using DFS – dynamic frequency selection).
  • Sometimes, there are simply too many interfering radios on the same channel or these radios are too ‘close-by’ on adjacent channels (an adjacent channel’s transmitter power spills over into a channel and affect the noise and interference levels, thereby affecting that radio’s ability to operate with the highest-order modulation based upon the modulation’s required E_b/N_o).
  • Sometimes, IP radio systems are affected by other radio systems in the same band that adjust channel access parameters to change ‘airtime fairness’ or implement ‘airtime unfairness’ (electing to use such methods can be to the detriment of other IP radio systems).
  • the possibility of retransmissions if the original transmission was not successful
• the use of TDD instead of FDD (for TDD, the same channel frequency is shared for both transmit and receive – furthermore, if the traffic flow is unbalanced and variable, there are resultant latency effects).
• the use of Layer 3 IP services can add additional delays e.g. NAT, Firewall rules
• Furthermore, ‘IP Radios’ will jump between MCS levels (or channels), attempting to mitigate effects of channel interference. Higher-order modulation methods are more sensitive to interference than more robust ‘lower-order modulations’ such as BPSK or QPSK.
  • If lower-order modulation must be used, either due to interference, broadcast or multicast traffic, there is less-efficient use of ‘air-time’ and hence higher and more-variable latency.
• Since IP radios share the frequency channel in license-free environments, as the number of channel users attempting to share the channel increases, the throughput will reduce and latency will increase (and become more variable).

Excessive and variable time delays and data transmission errors for voice packets transferred through backbone radios across a radio network results in the perception of poor-quality voice connections between user handsets.

For a voice communication network, the objective is to minimize voice packet errors, latency and jitter (also referred to as variation of latency) throughout the network.

The above list includes some of the ‘IP radio’ features that can affect the end-user’s perception of voice quality.  Bear in mind that QoS mechanisms (tagging of Ethernet packets) can assist by prioritizing the sending of some packets (e.g. ‘voice’) ahead of others (e.g. ‘best effort’ data).

Sagittar Microwave Radio Links – Africa, Middle East and SE Asia

Sagittar Microwave Radio Link Solutions

Email us at [email protected]

Or, please follow this link to ask questions and request additional information.