Potential Health Impacts of Gamma-Frequency Sound in Server Room

AHMED N. H. ALNUAIMY, RASHA HASHEM, MICHAEL JOHNSON

Department of Computer Techniques Engineering

Al-Rafidain University College

Baghdad | Palestine Street

IRAQ

Abstract: - Hearing degradation caused by an exposure to excessive noise is irreversible. Many of the other

relative hazards that can be developed from noise-induced hearing loss are impaired communication with

family and coworkers, social isolation, irritability, decreasing of self-esteem, anxiety, and loss of productivity.

Hearing impairment is a significant health hazard that is naturally occurs with aging. Tinnitus is a disorder in

hearing ability and can cause a ringing in the ear without a source for physical sound. More than 40 million

people in the United States are suffering from tinnitus disorder. Fourteen percent of adults are suffering from

chronic tinnitus, and 50% of normal adults with no clinically confirmed disorders in hearing ability experience

subtle tinnitus in a silent environment. An exposure to excessive noise and the process of natural aging of

people may increase the occurrence of Tinnitus. Tinnitus is a spontaneous auditory perception that is associated

with the continued activity of the gamma frequency band (30 Hz - 80Hz). Server Room can be considered as a

continuous source of gamma frequency. Server room running devices are generating a continuous noise that

most of its power is allocated in the band of gamma frequency.

Key-Words: - Noise, Server Room, OSHA, Gamma Frequency, Tinnitus.

Received: April 12, 2021. Revised: January 26, 2022. Accepted: February 24, 2022. Published: April 2, 2022.

1 Introduction

Large server rooms are considered as a new

workplace phenomenon and verify merit as an

emerging technology. Excessive noise exposure to

employees working with servers may be present in

data storage sites [1]. Corporate server rooms

provide a single location for residential computers

to support business goals, which are usually a small

temperature-controlled, secure room with a

minimum number of passengers. Maintaining a

small area allows greater access control (for

example, a high degree of security), and also puts

the servers in a more intense format.

The server is designed to deliver data to other

computers (clients) and process requests over the

Internet, or a local network [2]. Those rooms

contain multiple devices responsible for transmitting

data to and from these data center. These rooms

have different densities and dimensions depending

on the room’s architecture, brand of the server, the

server’s age, and racks height in addition to the

proximity of the units to each other. Strategic

placement of units can amplify sound in every

server room, as the sound pressure depends on the

number of servers and how well they are positioned

from each other. Proximity and density of servers

have the ability to produce excessive harmful noise.

Some operators may find the volume and frequency

uncomfortable, and excessive noise may be

dangerous to their well-being.

Sound parameters are pressure and frequency

that are measured in Newton/Square Meter (N/m2)

or Pascal (Pa) and (Hertz) respectively. Pressure and

frequency have health hazard on the users, but the

pressure of the sound is the only factor that is

regulated by OSHA (Occupational Safety and

Health Administration) [3]. Sound in server rooms

has a frequency resonant that may be offensive to

some people and can somehow cause health issue,

but unfortunately is not regulated by any of

governmental agency. OSHA mandated that sound

pressure levels should not exceed a certain level so

that they do not cause an occupational health issue

to employees. In this study, we evaluated and

investigated sound frequency as it may adversely

impact the users. For example, gamma frequency

(40 Hertz–100 Hertz) of the sound band has been

discovered to be uncomfortable or distracting for

some individuals [4]. The average hearing range of

the person is maintained from 20 to 20,000 Hertz.

The range of the frequency investigated in this study

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

can be reviewed in the Result section of this paper.

The objective of this research was to investigate if

the sound generated by the servers caused

discomfort for the users that maintain and use those

data center and also to analyze the characteristics of

noise.

The purpose of this study was to find out how

much did the servers generate of sound (power and

frequency) in a specific area, and then evaluate

whether the sound levels at gamma frequency need

to be reduced to avoid health hazard or issues

related to the productivity for the users. The

preferred way to address occupational exposure is

through engineering methods, preferably in the

design process. If the room or equipment can be

configured in a manner that eliminates or reduces

occupational exposures to the workers in the

workplace, then this could be the first approach by

ergonomic engineer supposing that it is

economically feasible. Engineering solutions (i.e.,

workplace design to reduce or avoid workers

exposure to risk condition) do not necessitate

administrative controls such as reduction of the

work hours of employees exposed to a certain

contaminant (in this case, sound), rotation of

employees in a specific job position, or mandate

annual medical test for hearing and compliance by

wearing equipment for safety purposes. The

preferred approach for reduction of noise impact

should always involve engineering solution first,

then administrative controls and finally resort

protective personal equipment.

The personal usage of protective equipment may

require the documentation for hearing capabilities of

each employee, and an investigation of the sound

reduction required amount that is defined by OSHA

29CFR1910.95 Hearing Conservation to ensure

compliance with regulation [3]. Annually, the cost

of conducting tests in order to determine employees’

hearing capabilities to choose convenient personal

equipment for hearing protectively can be expensive

in the long run.

Acoustics can be achieved or reduced in several

different ways depending on room configurations,

equipment placed in rooms, floor/wall construction

restrictions, and sound reduction materials. As an

engineering solution, noise encapsulation potentially

reduces the sound by isolating the device that emit

the sound from the user by enclosing it, completely.

Sound reduction methods may still allow the

equipment to be accessed by the user easily and

keep the sound below harmful levels. Sound

reduction involves the usage of appropriate personal

protective equipment (ear muffs, ear plugs, etc.) if

the pressure levels of the sound exceed the values

defined by OSHA. From engineering perspective,

noise cancellation is considered as another method

of sound reduction and it is more effective in work

environment. Noise cancellation method is

consisting of devices that use a technology of signal

processing to reduce the noise by capturing the

noise signal and then emitting an inverting version

of the sound wave, it is thereby canceling some of

the noise waves. Other forms of sound reduction are

accomplished by utilizing acoustic materials on

floor, ceilings, or walls, and that reduce the pressure

of the noise that emitted by the devices [5]. Personal

Protective Equipment (PPE) is typically employed

by organizations when engineering and

administrative solutions do not minimize sound to

an acceptable level to comply with federal

regulations [1]. Administrative solutions to protect

the hearing of employees include rotating

employees reducing employee exposure to noise.

These solutions create challenges in the scheduling

for organizations and are therefore not as desirable

as reducing or eliminating the sound exposure.

Lastly, the use of earmuffs, earplugs or any other

similar PPE may adversely impact the workers

ability to communicate while they are in a server

room and cause errors during the work due to poor

communication [5].

2 Tinnitus

Tinnitus is hearing impairment that is typically

associated with damage that occurs due to noise

trauma or chronically noise exposure [4]. Such

damage can hurt the central auditory system,

specifically the neural synchrony within the central

auditory system. These changes have been reported

in various studies conducted for animals and

humans and could be the cause of various

pathologies [4]. These damages have been reported

to be specifically tied to the gamma range (30 Hz –

80 Hz) of sound frequency [4]. Tinnitus has been

characterized as one of the common auditory

disorders in the population. The anatomical

substrates and disease associations still continue to

be defined. However, Semantic Dementia (SemD)

patients are frequently reporting Tinnitus as one of

the symptoms [6]. Therefore, it may be possible to

report significantly the potential onslaught or

prevalence of Tinnitus as a serious issue in SemD.

SemD evidence can support previous work

implicating of limbic network and a distributed

cortico-subcortical auditory in the pathogenesis of

these abnormal auditory percepts [6].

In general, Tinnitus is caused by peripheral and

central mechanisms such as peripheral injury, a

reorganization of central auditory pathways, or

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

anomalies in the limbic system, which produce a

sensory emotional content experience [7]. There are

many different hypotheses to explain the reason of

Tinnitus. Some evidence suggests that Tinnitus can

be a pathophysiology that involves damage either

central or peripheral pathway or can be both [8].

Another Hypothesis proposed that Tonotopic maps

could be the cause of tinnitus. Tonotopic maps are

recognized in the auditory cortex and leads to a

sensation of Tinnitus frequencies [7]. Previously, it

was a suggested that the source of chronic Tinnitus

are coming by a compromised limbic corticostriatal

circuit that leads to a disordered in evaluation of

Tinnitus sensation’s perceptual and causing

disturbance in the control of cognition in the

tinnitus. [8]. On the other hand, anomalies related

to Tinnitus are inter-correlated between primary and

limbic auditory and/or two limbic area and indicates

necessity of auditory-limbic interactions in Tinnitus

as shown in figure 1. Although, the role of limbic

contributions nature to Tinnitus is not yet verified

[8]. Auditory-Limbic, shown in figure 1, clarifies

the interaction in Tinnitus, where the sensory input

is originated subcortically then enters both auditory

and limbic circuits via Medial Geniculate Nucleus

(MGN).

Medial Dorsal Nucleus (MDN),

Ventral Pallidum (VP),

amygdala (amyg),

Auditory Cortex (AC),

Medial Geniculate Nucleus (MGN).

Fig 1. Auditory-limbic interaction in Tinnitus [8].

Normally, sensory signal is defined by the limbic

system as perceptually irrelevant such as transient

Tinnitus followed by noise exposure. Then block

unwanted signal to MGN by projections from the

ventromedial Prefrontal Cortex (vmPFC) to the

auditory thalamic reticular nucleus (TRN, red

pathway). Therefore, an unwanted signal can be

reduced in either circuit. In the case of tinnitus,

inactive vmPFC output can prevent tinnitus signal

hosting and cortical thalamic activity. The structures

of thalamus in blue, and amygdala is noted in

lavender, basal ganglia noted in green, and cortical

are noted in gray. Mapping the hubs of the cortical

in Tinnitus it has been reported that having major

group differences across global networks, especially

in the gamma frequency band [4].

Typically, Tinnitus is usually associated with

damage to the hearing system such as chronic

exposure to noise or noise trauma [4]. This damage

can lead to drastic changes at different levels of the

central auditory system. As a result, it boosted the

rate of automatic fire and nervous synchronization

within the central auditory system. These changes

have been reported in both animal and human

studies and may be caused by various diseases [4].

Another suggestion about the cause of chronic

Tinnitus mentioned that it may be caused by plastic

reorganization in the auditory cortex followed by

peripheral deafferentation. Based on this hypothesis,

the reorganization process usually causes hair cells

to be lost in the inner ear; Sensorineural Hear

Losing (SNHL) may in some cases result in

cochlear injury due to sound trauma (i.e. exposure to

noise with a certain level of frequency band or

associated atrophy Age for the hair cell).

Moreover, the corresponding frequency range

can cause the lesion thresholds to rise. Furthermore,

the adjacent frequencies became more amplified due

to the expansion of the central representation in a

vacant frequency band [9]. Indeed, some of the

basic findings of Positron Emission Tomography

(PET) studies show that the frequencies

corresponding to the perceived Tinnitus frequencies

lead to frequency expansion in the auditory cortex

[9]. While tinnitus is usually considered a

heterogeneous condition, most patients who have

suffered from tinnitus have reported a complaint of

a sense of auditory weakness. With regard to the

brain mechanism of tinnitus sufferers, most current

data show a very important gray matter that shrinks

in the subcallosal region as shown in figure 2 [7].

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

Fig 2. Gray-matter volume decreases in addition to

the changes throughout the whole brain [7].

Furthermore, it was found that there was an

expansion in the concentration of gray-matter in the

auditory thalamus of the Tinnitus group as shown in

figure 3 [7]. It is important to find that the structural

changes that associate with the region without

leaving the Tinnitus for different reasons like

activity in the artificial sub-region associated with

the unwanted auditory sensation that comes from

different amounts of disharmony symmetry,

especially in the area where the gray color is

shrinking the issue [7].

Fig 3. Gray-matter concentration increases [7].

Mainly, the role of the posterior thalamus and

subcallosal region in causing Tinnitus is combined

changes in both regions. In other words, these

changes seem to result in a tinnitus sensation.

Mulaw et al. (2005) A model suggested this:

1. Neural tinnitus activity is mainly in MGN and

results from reorganization after peripheral

hearing loss.

2. Inhibitory feedback from the subcallosal area

may help control nervous activity due to

Tinnitus.

3. Shrinkage of gray matter in the subcallosal

region may reduce these inhibitory reactions.

Because of this, people with peripheral hearing

loss may be at a health risk for tinnitus.

3 Results

A sound record was conducted in a workstation,

which located in an area where the operators could

perform paperwork or work on their computers with

tasks that were assigned to them. The recorded

signal time domain, frequency spectrum, and

spectrogram analysis are exhibited in figure 4-6

respectively. According to the time domain of the

sound (figure 4), sound pressure is bouncing around

0.05 Pa. However, acoustic (i.e., sound) pressure

level was plotted to show the noise pressure over

time. Refereeing to [1], the sound pressure level

(SPL) for the source of sound with pressure (p), is

defined as:

  





.. (1)

where pref= 20  10-6 Pa.

Therefore, the pressure of the sound at that

location could be exceeding the level of 60 dBA.

Therefore, it is not close to the maximum level of

the sound that is established by OSHA. Although

the noise level of the signal is not legally harmful,

due to the distance between the workstation and the

servers that causes a reduction in the SPL, it was

still representing a continuous distraction to the

workers in that server room.

A frequency spectrum analysis (figure 5) reveals

that the highest power of the signal is located in the

lower frequency band (i.e., 20–80 Hertz). At this

band, the power of the signal is ranging from -35 dB

to -60 dB. In figure 6, the signal power distribution

over the frequency is displayed in the time domain.

A frequency range of 0–120 Hz for the spectrogram

(figure 6) is evident in Figure 7. It shows that the

frequency band from 40–80 Hz has a level of power

that fluctuated between -50 dB to -60 dB. This range

is representing a gamma frequency range, which is

considered as the most harmful range in terms of

hearing impairment.

As a result, employees were exposed to a high-

power level of continuous gamma frequency sound

while spending the majority of their work shift in

the server room to perform daily tasks and monitor

the network. As mentioned earlier, long-term

exposure in this frequency range may cause

significant damage to the hearing system such as

chronic noise trauma [4].

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

Fig 4. Time domain for the sound pressure.

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

Fig 5. Frequency spectrum for the noise signal.

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

Fig 6. Frequency spectrum in time domain for the noise signal.

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

Fig 7. Spectrogram for noise signal.

4 Conclusion

In this paper, it was hypothesized that high sound

level at low frequency can cause a deleterious effect

on individuals who exposed to it for extended

periods. Some evidence supports this theory, and

this paper summarizes similar facts. The purpose

was to promote more investigative work in this area

and facilitate this discussion.

The impact of noise on job performance should

not rely on sound level only as parameter. It should

look up the frequency range of the noise too, as the

frequency can influence a person’s hearing ability

and even mental functionality. According to this

research effort, it is recommended that employees in

work environments similar to the server room or

data center to wear personal protective equipment

until an appropriate engineering solution can be

implemented to reduce noise levels.

Finally, it is recommended to collect more data

from inside server rooms with different

configurations, which help in investigation and

validation the impact of this sort of noise on human

performance in occupational environments.

References:

[1] Sultan, H., Katiyar, A., & Smruti , S. R., Noise

aware scheduling in data centers. Proceedings

of the 2016 International Conference on

Supercomputing. ACM., New York, NY:

ACM, 2016, pp. 13.

[2] Bryant, R. E., & O’Hallaron, D. R., Computer

systems: A programmer’s perspective. (3rd

ed.). Prentice Hall 4 Class Meetings Section

Time Days Classroom Date Range, 2015.

[3] OSHA. Occupational noise sxposure.

Retrieved February 2017, from

https://www.osha.gov/SLTC/noisehearingcons

ervation/, 2014.

[4] Schlee, W., Mueller, N., Hartmann, T., Keil, J.,

Lorenz, I., & Weisz, N., Mapping cortical hubs

in tinnitus. BMC Biology, 7(1), 80, 2009.

[5] Sharma, M. K., & Vig, R., Server noise: Health

hazard and its reduction using active noise

control. In 2014 Recent Advances in

Engineering and Computational Sciences

(RAECS), IEEE, March , 2014, pp. 1–5.

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022

[6] Mahoney, C. J., Rohrer, J. D., Goll, J. C., Fox,

N. C., Rossor, M. N., & Warren, J. D.

Structural neuroanatomy of tinnitus and

hypercusis in semantic dementia. Journal of

Neurology, Neurosurgery and Psychiatry,

82(11), November, 2011, pp. 1274–1278.

[7] Mühlau, M., Rauschecker, J. P., Oestreicher,

E., Gaser, C., Röttinger, M., Wohlschläger, A.

M., Simon, F., Etgen, T., Conrad, B., &

Sander, D., Structural brain changes in tinnitus.

Cerebral Cortex, 16(9), 2005, pp. 1283–1288.

[8] Leaver, A. M., Renier, L., Chevillet, M. A.,

Morgan, S., Kim, H. J., & Rauschecker, J. P.,

Dysregulation of limbic and auditory networks

in tinnitus. Neuron, 69(1), 2011, pp. 33–43.

[9] Rauschecker, J. P., Leaver, A. M., & Mühlau,

M., Tuning out the noise: Limbic-auditory

interactions in tinnitus. Neuron, 66(6), 2010,

pp. 819–826.

Contribution of individual authors to

the creation of a scientific article

(ghostwriting policy)

Ahmed N. H. Alnauimy carried out the data

collection and statistical analysis.

Michael Johnson assisted in data collection.

Rasha Hashem assisted in the Statistics.

Creative Commons Attribution License 4.0

(Attribution 4.0 International, CC BY 4.0)

This article is published under the terms of the Creative

Commons Attribution License 4.0

https://creativecommons.org/licenses/by/4.0/deed.en_US

WSEAS TRANSACTIONS on SIGNAL PROCESSING

DOI: 10.37394/232014.2022.18.7

Ahmed N. H. Alnuaimy,

Rasha Hashem, Michael Johnson

E-ISSN: 2224-3488

Volume 18, 2022