With Bluetooth™ initiatives placing bite marks on wireless communications, an associated problem that looms in the horizon in the form of EMI is attributed to electromagnetic emanations from high power ISM devices, such as microwave ovens. Since Bluetooth is largely conceived as an indoor, ad hoc mobile communication facility and, inasmuch as the microwave oven is an indispensable domestic gizmo, the question is how to resolve the compatibility issues on the persisting EMI on the Bluetooth interface caused by microwave leakage from the oven. Suggested in this article is a possible strategy wherein the microwave oven itself is made into a Bluetooth-enabled device, which would function as a master and perform the power control in other Bluetooth communication devices operating in the vicinity. Such devices thereby adjust their signal-to-interference ratio (whenever the oven is turned on) and could possibly remain immune from adverse effects of packet-losses due to the EMI under discussion.

P.S. Neelakanta, Jesada Sivaraks and Charoonkierti Thammakoranonta
Department of Electrical Engineering
Florida Atlantic University
Boca Raton, FL


Bluetooth technology is an interesting, topnotch technology that has been ushered in and enthusiastically promoted so as to enable devices of all kinds -- from laptops and cell phones, to personal digital assistants and household appliances -- the capability of communicating and interoperating with one another in the wireless medium. Taking a tour into the emerging vista of Bluetooth indicates the underlying concept of this technology as an enabling means towards a transparent wireless RF communication between electronic devices. This mode of communication can offer a set of advantages encompassing a viable voice/data access point, no cumbersome cabling and creation of an as-you-please ad hoc wireless network. Bluetooth technology is characterized by personal connectivity of enabled devices, spontaneous creation of disposable networks and low cost, low power indoor wireless transmissions of short-ranges supporting 721 kbps or 432.6 kbps over asynchronous and synchronous links, respectively.5

However, since the inception of Bluetooth, the hurdles on its technology have also been identified. Such hurdles refer to noisy radio environments in which the Bluetooth communication may take place. Specific compatibility issues of concern stem from the operation of Bluetooth in the proximity of high power microwave appliances, which use the same ISM band designated for the unlicensed spectrum used by the Bluetooth equipment as well.

The microwave oven is a typical high power, ISM band specified, indoor piece of equipment. The omnipresence of microwave ovens in the home and at work could pose considerable EMI influence on the operation of Bluetooth communications taking place in the vicinity; this EMI spilling out the inevitable (but permissible) level of microwave leakage (around 2.45 GHz) arising from the oven. Recent studies1,2 suggest that within the allocated ISM band, the emissions from modern microwave ovens could encroach on the working spectrum of nearby Bluetooth units by adversely influencing the associated frequency-hopping spread spectrum (FHSS) protocols. In particular, such influences could be significant on lower power Bluetooth systems (operating at 1 to 10 mW transmitter power). It has been observed that relevant operational impairment could occur even when the microwave oven is at a distance ten times further away from the communication units. This situation is still expected to be unfavorable even for higher power Bluetooth units if the oven is closer than tens of meters to the nearest Bluetooth-enabled device. Thus, the persistence of the problem is quite obvious. In essence, the microwave ovens, though a boon to homemakers, could prove to be a nuisance for Bluetooth units, which share the common ISM spectrum with microwave ovens. Now the question is how to resolve the compatibility issues between such sharing partners of the ISM band, namely, the Bluetooth and high power systems.

The indoor electromagnetic environment in which Bluetooth communication devices are to be deployed cannot be changed radically. There are millions of microwave ovens (and/or other ISM band high power domestic/industrial appliances) on the market, and an equal number of them have already proliferated households and are in use across the world. Their market and deployment cannot be altered to allow the use of Bluetooth in the absence of microwave ovens in the ambient. As such, a unique situation prevails -- people must live in the world of microwave ovens (present both domestically and within clustered office cubicles) and still make the best use of this new Bluetooth technology. As correctly stated by Buffler and Risman,1 "ISM and communications communities must, (therefore) work together to find some practical solution to the (associated) compatibility problems."

Motivated by the above suggestion, a strategy is presented to address the relevant compatibility problems. The proposal is as follows:

Why not make (on an ad hoc basis) all microwave ovens (to be manufactured and sold) Bluetooth-enabled so that they can be deployed as masters to control the transmitting power levels of all other Bluetooth communication units in the scattered ad hoc network operating in the vicinity?

Presented here are details on the methodology illustrating the proposal. For this purpose, the EMI environment of an operating microwave oven is first described and analyzed so as to identify the possible modes of interference on the Bluetooth equipment; hence, relevant avenues for mitigation strategies are searched and discussed.

The concept of Bluetooth as originally conceived and evolved is clear: It allows electronic gadgets and gizmos to talk to each other without wire connectors. In essence, Bluetooth operations are conceived to provide pragmatic alleys of communication access and permit information exchange within the web of information superhighways involving an interplay of a variety of electronic devices/units in a confined locale.

As indicated before, though wisely chosen to operate in the freebee spectral band for its universal applications at 2.4 GHz, the Bluetooth technology must face unfavorable alterations of its operating environment caused by the other electronic systems in operation which use the same ISM frequency. In general, such an electronic appliance could be a low power communication unit (for example, a garage door opener) or a high power system like a microwave oven emitting unintentional electromagnetic leakage.

Specifically, the high power appliances may operate at nominal microwave power ratings (in the order of 1 kW), but the expected microwave leakage from them could be high enough to cause interference on Bluetooth operation. This has been confirmed via measurements.

When microwave ovens came to the market, the original concern about such leakages was mostly directed to limit the specific absorption rate (SAR) dosage, which could otherwise prove harmful to living systems. Health and/or safety regulation bodies around the world prescribed leakage levels permissible in the proximity of microwave ovens so that manufacturers would comply with these regulatory constraints.

However, such a permissible SAR is not adequate to ensure EMI-free operation of Bluetooth-enabled equipment. This is because Bluetooth is a battery-operated low power system, and as the active (signal) power levels at the receiver side of a Bluetooth unit become comparable to microwave oven leakage, communication operation could be severely impaired.

Typical considerations and spectral emission characteristics of simple microwave ovens (holding a nominal food load) can be enumerated as follows: Microwave ovens may legally emit certain levels of leakage in the ISM band subject to the limits set by SAR-specified international safety standards. During the on-off transient states, the leakage from the oven may pose a spectral spread of approximately 20 MHz centered at 2.45 GHz. During such on-off transient states, the perceived leakage may increase by 5 to 15 percent of the nominal leakage level emitted when the oven is in steady-state on conditions.

The emitted microwave leakage level could potentially impair and affect the performance of direct sequence spread spectrum (DSSS)/
FHSS protocols of wireless communications. The emission characteristics (amplitude and spectrum) of the leakage may change radically for different types of food loads placed in the oven, and the leakage levels and spectrum of microwave energy from the oven could also be different for different versions of the ovens (such as a stirrer-type or turntable-type).

Mostly, the studies on microwave leakage have been performed with simple water loads, and relevant measurements address the extent of leakage power. This measured data can be directly related to the SAR value as required by health and safety regulations. But these measured artifacts have not been intended to link them as EMI influences on the Bluetooth equipment. (This is because Bluetooth did not exist and interference of relevance was not addressed. Also, the FCC allowed the general use of the ISM frequencies by any potential user as long as the new band inhabitant accepted interference that was caused by the operation of other authorized ISM equipment.)

The scenario under discussion is different. It is a conflict in the battlefield of the ISM band assigned for tranquil enjoyment by two parties -- the microwave ovens, which show their dominance with their privileged emission levels, and the low power Bluetooth units that seek to coexist with them.

In order to resolve this battle, it is necessary to understand more precisely the EMI due to microwave leakage from the ovens. In the words of Buffler and Risman1 "the allocated ISM band spectral characteristics of modern microwave ovens is far more complex in real life than that which is measured with simple water loads according to standardized methods for immunity and emission testing outside the band."

True. But how complex are relevant characteristics, which would directly affect Bluetooth performance? Any mitigatory effort would warrant a straight answer as well as elucidating an imminent data base on appropriate factors encroaching the domain of Bluetooth.

First of all, as indicated earlier, the leakage measurements on microwave ovens largely provide the net power level distribution profile around the oven structure and, invariably, no polarization characteristics of the emitted radiation are furnished. Such information on polarization of the emitted microwaves could, on the other hand, be important and can be profitably considered in the mitigation trails with regard to Bluetooth communication deployments.

Secondly, the spread or window of leakage spectrum should be critically reviewed to determine possible interference effects over the pseudo-random, frequency-hopped, gaussian-filtered channels adopted for Bluetooth transmissions.

The third consideration would refer to bouncing of leakage emissions from the ovens as a result of reflections in an indoor environment. Relevant implications on the performance of Bluetooth units operating in the same indoor locations are crucial while looking into any possible mitigations.

Consistent with these considerations (plus the available details in the literature on measured data on power levels and spectral characteristics pertinent to transient and steady-state microwave leakage from the ovens), avenues for mitigations compatible for the protocol structure of the Bluetooth could be sought.

Measurements on the missing polarization details of the leakage from a microwave oven were conducted. A standard 800 W microwave oven (which has been in use for over a few years), was adopted as the device under test (DUT). The unit chosen has a single see-through window and a gasketed front opening. A simple water load (in a coffee mug) was used.

A diode-detector mounted on a fixture with the ability to sense the E-field in the vertical, horizontal or 45° cross-plane was facilitated. It could be positioned along three-dimensional locations corresponding to the front, top, side and rear aspects of the oven. A grid structure made of parallel conducting wires was used as the Fresnel filter to select vertical, horizontal or cross-polarized fields. For example, when this filter is kept with the wire grids y-directed, it would allow only the y-directed (E-field) on the other side of the grid where a (y-directed) diode probe measures this field. That is, the measured field at the probe is VV (vertical-vertical) polarized. Likewise, HH (horizontal-horizontal) and cross-polarized components were measured at discrete locations around the oven. The measurement locales in front of the oven, for example, are shown in Figure 1.

Using the water load, the oven was turned on for a period of three minutes (30 seconds for each measurement position) and the leakage level (of appropriate polarization) was recorded on a chart recorder. A sample set of measured data is shown in Figure 2.

A set of repeated measurements performed as described and an ensemble of data compiled thereof indicate that the electromagnetic field of the emitted leakage is largely vertically polarized around the oven. Furthermore, the measured results indicate the maximum leakage is perceived in front of the oven through the see-through window and door gaskets.

POSSIBLE MITIGATORY STRATEGIES

Using the available information in the literature on microwave oven leakage characteristics, plus the additional details ascertained as described, some possible ways of mitigating EMI relevant to simultaneous operation of Bluetooth units and microwave ovens when placed in the same geographical locations were proposed. Hence, comparing various options, a concrete scheme of mitigatory effort is identified.

The microwave oven can be made better-shielded through more stringent gasketing and absorbing see-through windows, thereby constricting the leakage level to much less than the value admissible through health and safety standards. However, the relevant increase in cost of the oven should not be overlooked. This is a prevention method whereby the leakage would not introduce an adverse signal-to-interference ratio on the Bluetooth systems operating close by.

The leakage as determined from the measured data is largely vertically polarized. It leads to a conceptual design whereby the Bluetooth transceiver system is made with polarization diversity capability, so as to reject the mainly vertically polarized oven leakage. Again, this method could be limited because of the portability of Bluetooth-enabled devices, which can restrict polarization diversity-based operations.

A third option indicated here as a versatile and viable technique is to make the microwave oven Bluetooth-enabled and allow it to be a partner in the Bluetooth piconet. By making the microwave oven itself a Bluetooth-enabled device, the oven is enabled to alert the nearby operating Bluetooth units to boost their transmitted power whenever the oven is switched on.

BLUETOOTH-ENABLED MICROWAVE OVENS -- A FOE ENTICED TO BE A FRIEND

Among the three options indicated, the provision of excess EMI shielding on the oven could increase the cost of the oven prohibitively, the volume and weight of the oven may increase, or it may have some impact on the aesthetic look of the ovens.

The second option is feasible; however, polarization diversity reception considerations on Bluetooth transmissions must be explored more extensively consistent with the available modules and standards.

The third option appears to be logical and implementable. With the decreasing cost on Bluetooth modules, making microwave ovens Bluetooth-enabled may not adversely enhance the price of the ovens in the market. The relevant strategy can be described with a brief overview of the operational aspects of Bluetooth.

Considering the state-of-the-art profile of Bluetooth technology, shown in Figure 3 are typical piconet topologies of networked Bluetooth devices.

To illustrate the concept of the mitigation under discussion, a simple scatternet configuration of the type shown in Figure 4 is considered wherein the microwave oven is assumed to be Bluetooth-enabled and is intended to be a master (client) controlling the power levels of transmissions associated with Bluetooth other master-slave communications devices in the locale.

It is assumed that all the Bluetooth devices (including the one installed on the microwave device) in the scatternet are familiar with the presence of the other a priori. Should a new Bluetooth pair (master/slave) enter the scatternet (which is allowed as per existing Bluetooth specifications), certain additional features will be required to be included in the network.

Considering the displayed scatternet, there are two states of ambient EMI: Whenever the microwave oven is on, there is a chance of EMI encroaching the other Bluetooth pairs already in communication. When the oven is off, the ambient is assumed to be EMI-free. It is proposed that, whenever the oven is on, the relevant state could be sensed by the Bluetooth device mounted on the microwave oven and the possibility of EMI is conveyed by this device (as an alert) to other Bluetooth devices (in communications sessions) so that those devices can boost their transmitted power levels in order to realize a good carrier-to-interference ratio.

The protocol involved in this proposal is shown in Figure 5 where the Bluetooth-enabled microwave oven is indicated as a master (client) and one of the Bluetooth client/server pairs in a communication session is regarded as the server responding to any command from the master (oven). The command to be deliberated by the oven is two-fold: Whenever the oven (master unit) is on, the command tells the slave units to boost the transmitted power levels. When the oven (master unit) is off, the corresponding command would tell the slave units to return to a normal operating power level (so as to save battery life).

The flow chart of protocol is self-explanatory. First, there is a logical if-check on the master (oven) pertinent to the oven being on or off. If on, a physical link would be established with the server units (slave). This allows a set of primitives to go down the protocol stack and the slaves receive the alert information to increase their transmission power. This state would remain sustained as long as the oven is on.

Once the oven is turned off, however, the corresponding state would be conveyed to the slaves through the (already established) physical link (from the oven) so that the transmission power levels can be returned to normal values. Also, the master (oven) can excuse itself and get released from the scatternet. This scheme is fully consistent with the Bluetooth specifications and protocol stack4 shown in Figure 6.

POWER CONTROL METHODS

In order to implement the described strategy, the power control methods in Bluetooth that already exist in the marketed devices can be used without resorting to any new techniques. In general, such power control methods refer to either a monotonic step-by-step (discrete) up-down technique or a continuous power control through linear ranging of current-operated (amplifier) circuitry.

The power control command from the oven can be used for optimizing the power consumption and overall interference level. The devices equipped with power control capability can optimize the output power through a command in the link LMP (see the flow chart). Normally, this is done by measuring the receive signal strength indicator (RSSI) and reporting whether the power should be increased or decreased. The strategy indicated in the flow chart emulates the relevant RSSI condition via the command received from the master (oven) to implement the power boost or decrease conditions. On the hardware side, the power control adjusts the voltage regulator of the device appropriately or performs a linear analog control on a power amplifier.

CONCLUSION

The operation of Bluetooth devices and a microwave oven in the same locale could lead to EMI problems arising from the leakage of microwave energy from the oven. A plausible means of mitigating such EMI influences by making microwave ovens Bluetooth-enabled has been proposed. With a prevailing trend in the decreasing cost of Bluetooth devices, mounting such devices as a part of the microwave oven may not be considered expensive. On the other hand, it can provide a robust and effective coexistence of microwave ovens and Bluetooth-enabled devices. Relevant operational considerations, specification compatibility and implementation feasibilities are quite evident from the details presented in this article.

Besides Bluetooth technology, other wireless communications systems operating in 2.4 GHz, such as wireless local area networks (LAN) (IEEE 802.11), may also be prone to interference due to leakage from microwave ovens. Also, the wireless LAN configuration (that uses either direct sequence or frequency-hopping for CDMA purposes) could itself be a significant interference source for Bluetooth systems. In short, by forcing coexistence, the Bluetooth and wireless LAN devices may pose as interference sources to one another. Consequently, the performance of Bluetooth-enabled devices operating in the same environment with wireless LAN and/or microwave ovens is a topic of concern in wireless telecommunication that warrants additional study. *


References

1. C.R. Buffler and P.O. Risman, "Compatibility Issues Between Bluetooth and High Power Systems," Microwave Journal, Vol. 43, No. 7, July 2000, pp. 126­134.

2. A. Kamerman and N. Erkocevic, "Microwave Oven Interference on Wireless LANs Operating in the 2.4 GHz ISM Band," The 8th IEEE International Symposium on Personal, Indoor and Mobile Radio Communication (PIMRC'97), Vol. 3, 1997, pp. 1221­1227.

3. Bluetooth Special Interest Group (SIG), Specification of the Bluetooth System V1.0B Core, Vol. 1, December 1999.

4. R. Mettala, "Bluetooth Protocol Architecture Version 1.0," Bluetooth Whitepaper, August 1999.

5. "Performing Bluetooth RF Measurements Today," Application Note 1333, Hewlett Packard Co., October 1999.

6. N.J. Muller, Bluetooth Demystified, McGraw-Hill, New York, 2000.

 

MWJ37NEELAKANTA P.S. Neelakanta received his PhD degree in electrical engineering from the Indian Institute of Technology, Madras (India) in 1975. He has over 30 years of experience (research, teaching and industrial) over a wide spectrum of interests covering electromagnetics, microwaves, antennas, radar and telecommunications (wireline and wireless). He has published extensively, authoring over 130 research papers and four books, including A Textbook on ATM Telecommunications (CRC Press, 2000). Dr. Neelakanta is a professor of electrical engineering at Florida Atlantic University (FAU), Boca Raton, FL, and his current areas of research are ATM telecommunications, access technology, WLAN/Bluetooth and stealth target RCS. Neelakanta can be reached via e-mail at neelakan@fau.edu

MWJ37SIVARAKS J. Sivaraks received his MSEE degree from Oklahoma State University and BEng from King Mongkut's Institute of Technology, Thailand. He is currently due to complete his PhD in electrical engineering at FAU. He is on deputation from the International Telecommunication Department of the Telephone Organization of Thailand. He has extensive experience in telecommunication planning and related system analysis. His PhD work is on the system aspects of Bluetooth, WLAN and Mobile IP/CDPD vis-á-vis associated interference issues. Sivaraks can be reached via e-mail at jsivara@hotmail.com

MWJ37THAMMAKO C. Thammakoranonta received his BEng degree from Srinakharinwirot University, Thailand, and is due to complete his MSEE at FAU. He has worked on analysis and fault diagnosis of mobile communication systems and radio base stations. He also has experience in frequency planning and fault correction efforts in wireless technology. His current interest is in wireless communications, including Bluetooth, WLAN and cellular technology. Thammakoranonta can be reached via e-mail at oak1480@hotmail.com or ctha0910@fau.edu.