Mission Effective Energy Efficiency (mEEE)

• Mission Effective Energy Efficiency (mEEE) is a framework that measures energy use by calculating the ratio of mission-valid utility to total power consumption.
• It spans diverse domains such as wireless communications, robotics, and network control, ensuring outcomes meet deadlines, reliability, and QoS constraints.
• The approach integrates performance, delay, and reliability metrics so energy expenditure is translated into genuinely useful work under mission constraints.

Searching arXiv for the cited papers to ground the article in current preprint records. Mission Effective Energy Efficiency (mEEE) is a mission-aware energy-efficiency construct in which energy expenditure is evaluated against useful work that remains operationally valid under task constraints such as deadlines, reliability targets, statistical delay guarantees, coverage requirements, or successful completion of mission objectives. In the most explicit formulation, developed for fluid antenna systems under finite blocklength and dependability constraints, mEEE is the ratio of mission effective capacity to total average power, $mEEE(\Phi)=mEC(\Phi)/P_t(\Phi)$ (Muhammad et al., 26 Jul 2025). Closely related formulations appear across adaptive computing, ultra-reliable short-packet communications, massive-MIMO network control, satellite–UAV access, quadrotor path planning, cooperative UAV–UGV routing, and payload-aware multirotor flight, where the common principle is that conventional energy efficiency is insufficient unless the numerator is filtered by mission feasibility (Dasari et al., 2018, Shehab et al., 2018, Hoffmann et al., 2021, Yacef et al., 2020, Tran et al., 2024, Mondal et al., 29 Apr 2025, Patnaik et al., 6 Jan 2025).

1. Conceptual scope and domain-specific meanings

Across the cited literature, mEEE is not a single domain-invariant formula; it is a family of ratios in which the numerator represents mission-valid utility and the denominator represents mission energy or power. What changes across domains is the precise definition of utility and the constraint set under which the utility is admissible.

Domain	Mission-valid numerator	Energy denominator
Adaptive computing	On-time feasible approximate task utility $U$	Compute and communication energy $E_{total}$
Finite-blocklength wireless	Effective capacity or mission effective capacity	Average consumed power
Massive-MIMO BS switching	$c_{50}$ only if served-UE constraint holds	Average BS power
Quadrotor mission control	Successful, timely, safe mission score $M$	Mission energy $E$
UAV–UGV routing	Completed tasks, optionally time-normalized	Total UAV, UGV, and charging energy
Multirotor forward flight	Specific energy-per-meter objective $\varepsilon$	Embedded in $P_{total}/(vm)$

This suggests that mEEE is best understood as a constrained utility-per-energy principle rather than a single canonical metric. A recurring structural feature is an indicator, reliability term, or feasibility constraint that suppresses nominal throughput or utility when mission conditions are violated. In that sense, mEEE generalizes ordinary energy efficiency in the same way that effective capacity generalizes Shannon throughput under delay constraints, or mission computability generalizes polynomial-time computability under fielded resource limits (Shehab et al., 2018, Dasari et al., 2018).

2. Mission computability as the abstract foundation

The most general formal substrate for mEEE in the supplied corpus is the mission-computability framework of adaptive computing systems (Dasari et al., 2018). That work defines deterministic polynomial time as

$P=\bigcup_{k\ge 0} Time(n^k),$

and introduces a mission class $M\subseteq P$ for problems whose instances admit feasible approximate solutions within mission time and resource limits. Using the paper’s notation, for a mission-focused relation $U$0 with $U$1,

$U$2

with approximation quality constrained by

$U$3

Mission computability further requires platform-dependent runtime to satisfy

$U$4

while analogous bounds may also apply to energy, memory, bandwidth, and thermal limits.

Within this framework, polynomial-time solvability is not enough. A language may lie in $U$5 yet fail to be mission-computable when any of the operative inequalities are violated, including $U$6 and $U$7, together with communication constraints such as bandwidth and link quality. The paper’s scheduling model makes this concrete by optimizing assignment variables $U$8 under one-image-per-job, at-most-once, and deadline-feasibility constraints, with extensions to distributed variables $U$9 that can include “transfer times, power used, or even the amount of RAM used” (Dasari et al., 2018).

A natural mEEE interpretation consistent with this formulation is mission utility per unit energy. The synthesis attached to the paper defines

$E_{total}$0

where $E_{total}$1 marks deadline satisfaction and $E_{total}$2 encodes approximation quality, and then proposes

$E_{total}$3

Because the original paper does not explicitly name mEEE, this should be read as an inferred metric grounded in its insistence on approximate solutions, mission deadlines, and resource-aware scheduling. Its significance is conceptual: energy efficiency becomes mission-effective only when the system converts energy into computations that remain useful before the mission window closes (Dasari et al., 2018).

3. Communication-theoretic formulations under finite blocklength, delay, and dependability

The communication-theoretic literature makes the distinction between plain energy efficiency and mEEE especially precise. In ultra-reliable finite-blocklength wireless links, effective capacity is defined as

$E_{total}$4

with

$E_{total}$5

where $E_{total}$6 is the finite-blocklength rate, $E_{total}$7 is the delay exponent, and $E_{total}$8 is Rayleigh fading power (Shehab et al., 2018). Under the linear power model $E_{total}$9, effective energy efficiency becomes

$c_{50}$0

The same work shows that buffer-aware modeling changes the denominator to

$c_{50}$1

and that accounting for empty-buffer probability jointly enhances effective capacity and effective energy efficiency. It also reports that Shannon’s asymptotic model underestimates the energy-efficiency-optimal power by about $c_{50}$2–$c_{50}$3 dB relative to the finite-blocklength model, which is mission-relevant when power budgets are tight (Shehab et al., 2018).

The fluid-antenna dependability framework turns this mission mapping into an explicit mEEE definition (Muhammad et al., 26 Jul 2025). There,

$c_{50}$4

with mission effective capacity

$c_{50}$5

and mission reliability

$c_{50}$6

The failure rate $c_{50}$7 is derived from second-order fading statistics—level-crossing rate and average fade duration—evaluated at the finite-blocklength decoding threshold. The corresponding power model is

$c_{50}$8

which explicitly captures bursty arrivals, idle periods, and non-empty-buffer probability surrogates. Optimization is posed as a non-convex fractional problem solved by a modified Dinkelbach method with embedded line search (Muhammad et al., 26 Jul 2025).

These two lines of work establish the core wireless interpretation of mEEE. First, the numerator is not raw throughput but delay-feasible service, and in the FAS case it is further gated by failure-free mission operation over a duration $c_{50}$9. Second, the denominator must model realistic power states, including idle, active, and buffer-dependent behavior. The resulting metric is therefore closer to “mission-valid bits per joule” than to generic link-layer EE.

4. Network infrastructure formulations: constrained service utility per power

In network control, mEEE commonly appears as constrained system-level throughput per infrastructure power. In massive-MIMO HetNets with switchable base stations, energy efficiency is defined per state $M$0 and action $M$1 as

$M$2

where $M$3 is the median downlink user bitrate and $M$4 is the average system power (Hoffmann et al., 2021). The mission constraint is that the number of served UEs must not fall below the all-BSs-active baseline:

$M$5

Accordingly, the reward is

$M$6

which is equivalent to the constrained form

$M$7

The proposed REM-plus-RL system uses contextual-bandit updates, analytical action space reduction, and a REM-based exploration algorithm. In the reported Madrid Grid scenario, it delivers a 70% EE gain over an analytical heuristic, while action filtering and REM-EA reduce convergence time by 60% and 83%, respectively, relative to comparison exploration baselines (Hoffmann et al., 2021).

A related but distinct formulation appears in integrated satellite–terrestrial UAV-enabled cell-free massive MIMO (Tran et al., 2024). There, the closed-form UAV-layer energy efficiency is

$M$8

subject to per-UAV power budgets and per-user QoS constraints

$M$9

The problem is solved with a successive convex approximation procedure that linearizes the non-convex perspective and SINR constraints. The study reports that “a few tens” of UAV access points transmitting with fine-tuned power are sufficient to empower the service of satellite networks and significantly increase spectral efficiency (Tran et al., 2024).

In both cases, mission effectiveness is enforced by service constraints before energy efficiency is evaluated. The difference is architectural. The massive-MIMO switching work enforces feasibility by zeroing the reward when served-UE coverage degrades, whereas the satellite–UAV framework embeds mission feasibility directly as QoS constraints in the optimization problem. Both are mEEE formulations in substance: utility counts only if the network remains mission-feasible.

5. Mobile robotic and aerial mission formulations

In robotics and aerial control, mEEE is often tied to physically successful mission execution rather than to communication service alone. For wind-aware quadrotor path planning, the energy-optimal controller minimizes

$E$0

under dynamic constraints, actuator bounds, and fixed terminal conditions (Yacef et al., 2020). A mission-level score consistent with the study is

$E$1

where success, timeliness, and quality/risk remain near unity in the reported experiments, yielding

$E$2

Under identical mission time and safety constraints, the optimal trajectory consumes $E$3 kJ versus $E$4 kJ for the adaptive controller in the no-wind case, a 67.24% energy saving and an $E$5 ratio improvement of about $E$6. With deterministic wind, the adaptive controller consumes $E$7 kJ and the optimal solution about $E$8 kJ, corresponding to about 44.0% lower energy and an $E$9 ratio improvement of about $\varepsilon$0 (Yacef et al., 2020).

The cooperative UAV–UGV routing framework is organized around makespan minimization rather than an explicit energy-efficiency objective, but its synthesis defines mission-centric energy metrics directly from the solved routes (Mondal et al., 29 Apr 2025). UAV segment energy is modeled as

$\varepsilon$1

with fixed-speed UAV power

$\varepsilon$2

and total energy

$\varepsilon$3

Two proposed metrics are

$\varepsilon$4

and

$\varepsilon$5

These are inferred rather than explicit in the original objective, but they are directly consistent with the model outputs. The reported routing gains are substantial: in the U45G15 case with 4 UAVs and 2 UGVs, DRL(10240) attains $\varepsilon$6 min versus $\varepsilon$7 min for GLS, $\varepsilon$8 min for TS, and $\varepsilon$9 min for SA; DRL(1024) remains within 2–3% of DRL(10240) while being 7–10× faster at inference (Mondal et al., 29 Apr 2025).

A more reduced aerodynamic interpretation appears in payload-aware multirotor forward flight (Patnaik et al., 6 Jan 2025). There the mission-efficiency quantity is the specific energy per meter

$P_{total}/(vm)$0

with maximum range

$P_{total}/(vm)$1

The paper proves that under the energy-optimal forward velocity, the optimal pitch angle is mass-invariant and the optimum specific energy is constant across vehicle mass:

$P_{total}/(vm)$2

For the tested octorotor in steady, level, no-wind flight over $P_{total}/(vm)$3 kg, the common optimum is reported near $P_{total}/(vm)$4 J/(m·kg), the optimal pitch near $P_{total}/(vm)$5, and a consistent empirical fit of the optimal speed is $P_{total}/(vm)$6 m/s (Patnaik et al., 6 Jan 2025). In this setting, mEEE is not mission utility per joule but the inverse problem of minimizing energy per payload-distance-normalized transport under the appropriate operating point.

6. Algorithms, trade-offs, and unresolved issues

The algorithmic machinery behind mEEE is heterogeneous because the constraint sets are heterogeneous. Adaptive mission computing uses integer programming and constraint-aware distributed scheduling with variables $P_{total}/(vm)$7 and $P_{total}/(vm)$8, together with approximate algorithms and early stopping under bounded mission time (Dasari et al., 2018). Finite-blocklength wireless optimization relies on closed-form approximations for effective energy efficiency and root finding for the power optimum, while the FAS formulation elevates the problem to non-convex fractional programming solved by a modified Dinkelbach method with line search (Shehab et al., 2018, Muhammad et al., 26 Jul 2025). Massive-MIMO base-station control uses contextual-bandit RL, analytical action filtering, and REM-guided exploration (Hoffmann et al., 2021). Integrated satellite–UAV cell-free massive MIMO uses successive convex approximation (Tran et al., 2024). Wind-aware quadrotor planning employs hp-adaptive Gaussian quadrature collocation via GPOPS-II with IPOPT (Yacef et al., 2020). Cooperative UAV–UGV routing uses an encoder–decoder transformer trained with REINFORCE and a greedy baseline, together with sortie-wise agent switching (Mondal et al., 29 Apr 2025).

Several cross-domain trade-offs recur. Accuracy versus timeliness is explicit in mission computability, where approximate solutions $P_{total}/(vm)$9 become mission-valid when they satisfy $P=\bigcup_{k\ge 0} Time(n^k),$0 before $P=\bigcup_{k\ge 0} Time(n^k),$1 (Dasari et al., 2018). Reliability versus energy is central to finite-blocklength and FAS systems, where tighter $P=\bigcup_{k\ge 0} Time(n^k),$2, larger $P=\bigcup_{k\ge 0} Time(n^k),$3, or higher mission-reliability targets reduce mEEE unless compensated by SNR, diversity, or delay relaxation (Shehab et al., 2018, Muhammad et al., 26 Jul 2025). Coverage or fairness versus power dominates network infrastructure settings, where switching off resources improves energy performance only if coverage or QoS constraints remain intact (Hoffmann et al., 2021, Tran et al., 2024). Time versus energy dominates vehicle planning, where faster completion can improve time-normalized mEEE even when absolute energy rises, but only so long as success, safety, and recharge feasibility are preserved (Yacef et al., 2020, Mondal et al., 29 Apr 2025).

A persistent misconception is that mEEE is just ordinary energy efficiency with a different label. The literature does not support that view. In the finite-blocklength setting, Shannon-based EE can misestimate the optimum transmit power (Shehab et al., 2018). In massive-MIMO switching, an action with high nominal $P=\bigcup_{k\ge 0} Time(n^k),$4 is worthless if it serves fewer users than the all-on baseline, because the reward collapses to zero (Hoffmann et al., 2021). In wind-aware UAV control, energy savings count as mEEE gains only because terminal constraints, timing, and safety are held fixed (Yacef et al., 2020). In several domains, moreover, the term itself is still partly inferential rather than standardized: the adaptive-computing, satellite–UAV, UAV–UGV, and payload-aware flight papers provide mission-consistent constructions, but not a universally shared taxonomy (Dasari et al., 2018, Tran et al., 2024, Mondal et al., 29 Apr 2025, Patnaik et al., 6 Jan 2025).

The open problems identified across the corpus are correspondingly structural. Adaptive computing still lacks explicit, calibrated energy models and formal mission-utility mappings (Dasari et al., 2018). Wind-aware UAV control omits stochastic wind uncertainty and higher-fidelity aerodynamic drag modeling (Yacef et al., 2020). Massive-MIMO switching does not yet price switching cost or hardware wear in the reward (Hoffmann et al., 2021). Satellite–UAV EE omits satellite power in the denominator and does not yet embed latency or reliability constraints (Tran et al., 2024). Cooperative UAV–UGV routing still assumes deterministic travel and charging models and would need stochastic energy, partial-charge, and multi-charger extensions for a fully mature mEEE theory (Mondal et al., 29 Apr 2025). Taken together, these gaps indicate that mEEE is already a coherent organizing principle, but not yet a universally closed formalism.

In aggregate, the literature converges on a precise interpretation: mEEE is the energy efficiency of useful work after mission admissibility has been enforced. Whether the admissibility filter is approximation quality, effective capacity under delay, mission reliability over a finite horizon, served-UE preservation, QoS-constrained throughput, successful transit, completed tasks, or minimum specific transport energy, the central idea is the same. Energy becomes mission-effective only when it is converted into outcomes that remain valid under the mission’s operational envelope (Muhammad et al., 26 Jul 2025, Shehab et al., 2018, Dasari et al., 2018).